id	sid	tid	token	lemma	pos
work_hea3gijzhnc2rl4sxgyj6pvghm	1	1	What	what	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	1	2	Makes	make	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	1	3	Writing	write	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	1	4	Great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	1	5	?	?	.
work_hea3gijzhnc2rl4sxgyj6pvghm	2	1	First	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	2	2	Experiments	Experiments	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	2	3	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	4	Article	Article	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	5	Quality	Quality	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	6	Prediction	Prediction	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	7	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	2	9	Science	Science	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	10	Journalism	Journalism	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	11	Domain	Domain	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	12	Annie	Annie	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	13	Louis	Louis	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	14	University	University	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	16	Pennsylvania	Pennsylvania	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	17	Philadelphia	Philadelphia	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	2	19	PA	pa	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	20	19104	19104	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	2	21	lannie@seas.upenn.edu	lannie@seas.upenn.edu	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	22	Ani	Ani	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	23	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	24	University	University	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	26	Pennsylvania	Pennsylvania	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	27	Philadelphia	Philadelphia	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	2	29	PA	PA	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	30	19104	19104	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	2	31	nenkova@seas.upenn.edu	nenkova@seas.upenn.edu	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	2	32	Abstract	Abstract	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	2	33	Great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	2	34	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	2	35	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	2	36	rare	rare	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	2	37	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	2	38	highly	highly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	2	39	admired	admired	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	2	40	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	3	1	Readers	reader	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	3	2	seek	seek	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	3	3	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	3	4	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	3	5	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	3	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	3	7	beautifully	beautifully	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	3	8	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	3	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	3	10	informative	informative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	3	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	3	12	entertaining	entertaining	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	3	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	4	1	Yet	yet	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	4	2	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	4	4	access	access	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	5	technologies	technology	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	4	6	lack	lack	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	4	7	capabil-	capabil-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	4	8	ities	itie	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	4	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	10	predicting	predict	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	4	11	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	12	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	13	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	14	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	4	15	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	4	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	5	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	2	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	5	3	paper	paper	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	5	5	present	present	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	5	6	first	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	5	7	experiments	experiment	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	5	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	9	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	10	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	11	prediction	prediction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	5	14	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	15	jour-	jour-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	5	16	nalism	nalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	17	domain	domain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	5	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	6	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	6	2	introduce	introduce	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	6	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	6	4	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	6	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	6	7	pieces	piece	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	6	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	9	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	10	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	6	12	along	along	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	13	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	14	typical	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	6	15	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	6	16	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	6	18	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	6	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	7	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	7	2	imple-	imple-	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	7	3	ment	ment	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	4	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	7	6	capture	capture	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	7	7	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	7	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	9	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	10	writ-	writ-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	7	11	ing	ing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	7	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	7	13	including	include	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	7	14	surprising	surprising	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	7	16	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	7	18	emotional	emotional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	19	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	7	21	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	7	22	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	7	23	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	24	general	general	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	25	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	26	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	27	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	28	discourse	discourse	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	7	29	organization	organization	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	30	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	7	31	sentence	sentence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	32	structure	structure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	7	33	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	8	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	8	2	show	show	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	8	3	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	8	5	distinction	distinction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	6	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	7	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	8	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	8	9	typical	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	8	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	8	11	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	8	12	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	8	13	detected	detect	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	14	fairly	fairly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	8	15	ac-	ac-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	8	16	curately	curately	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	8	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	8	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	8	19	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	8	21	entire	entire	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	8	22	spectrum	spectrum	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	23	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	24	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	8	25	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	8	26	contribute	contribute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	8	27	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	8	29	distinction	distinction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	8	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	9	1	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	9	2	Introduction	introduction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	3	Measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	9	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	5	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	6	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	7	would	would	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	9	8	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	9	9	hugely	hugely	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	9	10	bene-	bene-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	9	11	ficial	ficial	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	9	12	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	13	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	14	retrieval	retrieval	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	9	16	recommendation	recommendation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	9	17	systems	system	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	9	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	10	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	2	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	3	paper	paper	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	10	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	6	describe	describe	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	8	dataset	dataset	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	10	New	New	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	11	York	York	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	12	Times	Times	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	13	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	14	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	15	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	10	16	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	17	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	18	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	10	19	categorized	categorize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	20	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	21	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	22	differences	difference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	10	23	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	10	24	present	present	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	10	25	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	26	system	system	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	27	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	28	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	10	29	automatically	automatically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	10	30	make	make	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	10	31	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	10	32	distinction	distinction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	10	33	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	11	1	Science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	11	2	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	11	3	conveys	convey	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	11	4	complex	complex	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	11	5	scientific	scientific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	11	6	ideas	idea	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	11	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	11	8	entertaining	entertain	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	11	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	11	10	educating	educate	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	11	11	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	11	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	11	13	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	11	14	time	time	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	11	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	12	1	Consider	consider	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	12	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	12	3	following	follow	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	12	4	opening	opening	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	12	7	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	12	8	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	9	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	10	David	David	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	12	11	Quammen	Quammen	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	12	12	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	13	Harper	Harper	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	12	14	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	12	15	magazine	magazine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	16	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	12	17	One	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	12	18	morning	morning	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	19	early	early	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	12	20	last	last	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	12	21	winter	winter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	22	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	12	23	small	small	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	12	24	item	item	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	25	appeared	appear	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	12	26	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	27	my	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	12	28	local	local	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	12	29	newspaper	newspaper	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	30	announcing	announce	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	12	31	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	12	32	birth	birth	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	33	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	34	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	12	35	extraordi-	extraordi-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	12	36	nary	nary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	12	37	animal	animal	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	12	38	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	13	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	13	2	team	team	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	4	researchers	researcher	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	13	5	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	6	Texas	Texas	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	13	7	A&M	A&M	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	13	8	Uni-	Uni-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	13	9	versity	versity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	10	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	13	11	succeeded	succeed	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	13	cloning	clone	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	13	14	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	13	15	whitetail	whitetail	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	13	16	deer	deer	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	13	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	14	1	Never	never	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	14	2	done	do	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	14	3	before	before	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	14	4	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	15	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	15	2	fawn	fawn	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	15	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	15	4	known	know	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	15	5	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	15	6	Dewey	Dewey	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	15	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	15	8	was	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	15	9	developing	develop	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	15	10	normally	normally	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	15	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	15	12	seemed	seem	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	15	13	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	15	14	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	15	15	healthy	healthy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	15	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	16	1	He	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	16	2	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	16	3	no	no	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	16	4	mother	mother	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	16	6	just	just	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	16	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	16	8	surrogate	surrogate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	9	who	who	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	16	10	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	16	11	carried	carry	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	12	his	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	16	13	fetus	fetus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	14	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	15	term	term	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	16	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	17	1	He	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	17	2	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	17	3	no	no	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	17	4	father	father	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	17	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	17	6	just	just	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	17	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	17	8	“	"	``
work_hea3gijzhnc2rl4sxgyj6pvghm	17	9	donor	donor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	17	10	”	"	''
work_hea3gijzhnc2rl4sxgyj6pvghm	17	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	17	12	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	17	13	his	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	17	14	chromosomes	chromosome	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	17	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	18	1	He	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	18	2	was	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	18	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	18	4	genetic	genetic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	18	5	duplicate	duplicate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	18	8	certain	certain	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	18	9	trophy	trophy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	10	buck	buck	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	11	out	out	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	13	south	south	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	18	14	Texas	Texas	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	18	15	whose	whose	WP$
work_hea3gijzhnc2rl4sxgyj6pvghm	18	16	skin	skin	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	17	cells	cell	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	18	18	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	18	19	been	be	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	20	cultured	culture	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	21	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	22	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	18	23	laboratory	laboratory	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	18	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	19	1	One	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	3	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	4	cells	cell	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	19	5	furnished	furnish	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	7	nucleus	nucleus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	8	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	19	10	transplanted	transplant	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	19	12	rejiggered	rejiggere	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	19	14	became	become	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	16	DNA	DNA	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	19	17	core	core	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	18	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	19	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	20	egg	egg	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	21	cell	cell	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	19	23	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	24	became	become	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	25	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	26	embryo	embryo	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	27	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	19	28	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	19	29	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	30	time	time	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	19	31	be-	be-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	19	32	came	come	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	19	33	Dewey	Dewey	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	19	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	20	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	20	2	he	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	20	3	was	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	20	4	wildlife	wildlife	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	20	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	20	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	20	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	20	8	sense	sense	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	20	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	20	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	20	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	20	12	an-	an-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	20	13	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	20	14	sense	sense	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	20	15	elaborately	elaborately	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	20	16	synthetic	synthetic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	20	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	21	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	21	2	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	21	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	21	4	sort	sort	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	6	news	news	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	21	8	quirky	quirky	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	21	9	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	21	10	epochal	epochal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	21	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	21	12	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	21	13	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	21	14	cause	cause	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	21	15	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	21	16	person	person	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	17	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	18	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	21	19	mouth-	mouth-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	20	ful	ful	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	22	toast	toast	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	21	23	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	21	24	pause	pause	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	21	25	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	21	26	marvel	marvel	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	21	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	22	1	What	what	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	22	2	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	22	3	dumb	dumb	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	22	4	idea	idea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	22	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	22	6	I	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	22	7	marveled	marvel	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	22	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	23	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	23	2	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	4	clear	clear	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	23	6	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	23	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	23	8	organized	organize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	9	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	23	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	23	11	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	12	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	23	13	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	14	creative	creative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	15	use	use	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	17	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	23	19	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	23	20	clever	clever	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	21	story	story	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	22	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	23	23	like	like	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	24	explanation	explanation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	23	27	scientific	scientific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	23	28	con-	con-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	29	tribution	tribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	23	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	24	1	Such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	24	2	properties	property	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	24	3	make	make	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	24	4	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	5	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	6	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	24	7	attractive	attractive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	24	8	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	10	studying	study	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	24	11	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	12	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	24	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	25	1	Sci-	sci-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	25	2	ence	ence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	3	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	4	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	25	5	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	25	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	25	7	highly	highly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	25	8	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	25	9	domain	domain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	11	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	12	retrieval	retrieval	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	13	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	25	15	context	context	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	17	educational	educational	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	25	18	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	25	19	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	25	20	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	25	21	entertaining	entertaining	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	25	22	applications	application	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	25	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	26	1	Article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	26	2	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	26	3	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	26	4	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	26	5	hugely	hugely	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	26	6	benefit	benefit	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	26	7	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	26	8	systems	system	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	26	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	27	1	Prior	prior	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	2	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	3	indicates	indicate	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	4	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	5	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	6	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	27	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	8	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	9	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	10	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	11	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	27	12	successfully	successfully	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	27	13	predicted	predict	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	14	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	27	15	a	a	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	27	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	17	whether	whether	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	18	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	27	19	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	20	meets	meet	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	27	22	acceptable	acceptable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	23	standards	standard	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	27	24	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	25	spelling	spelling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	26	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	27	Brill	Brill	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	28	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	27	29	Moore	Moore	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	31	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	32	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	33	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	34	grammar	grammar	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	27	35	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	36	Tetreault	Tetreault	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	37	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	27	38	Chodorow	Chodorow	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	40	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	41	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	27	42	Rozovskaya	Rozovskaya	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	43	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	27	44	Roth	Roth	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	45	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	46	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	47	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	48	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	27	49	discourse	discourse	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	50	organization	organization	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	51	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	52	Barzilay	Barzilay	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	53	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	54	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	55	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	56	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	57	2002	2002	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	58	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	27	59	Lap-	Lap-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	60	ata	ata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	27	61	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	27	62	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	27	63	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	64	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	27	65	b	b	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	27	66	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	27	67	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	68	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	27	69	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	70	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	27	71	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	72	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	73	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	74	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	27	75	par-	par-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	76	ticular	ticular	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	27	77	user	user	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	27	78	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	28	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	2	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	28	4	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	28	6	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	7	recommen-	recommen-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	8	dation	dation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	9	systems	system	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	10	standardly	standardly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	28	11	represent	represent	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	12	user	user	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	13	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	14	us-	us-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	28	15	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	28	16	frequent	frequent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	28	17	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	18	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	19	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	20	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	28	22	user	user	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	23	’s	’s	,
work_hea3gijzhnc2rl4sxgyj6pvghm	28	24	history	history	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	25	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	28	26	retrieve	retrieve	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	28	27	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	28	28	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	29	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	30	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	28	31	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	28	32	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	33	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	28	34	Paz-	Paz-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	35	341	341	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	28	36	Transactions	Transactions	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	28	37	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	38	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	28	39	Association	Association	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	40	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	28	41	Computational	Computational	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	42	Linguistics	Linguistics	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	28	43	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	28	44	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	28	45	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	28	46	2013	2013	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	28	47	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	28	48	341–352	341–352	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	28	49	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	29	1	Action	Action	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	29	2	Editor	Editor	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	29	3	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	29	4	Mirella	Mirella	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	29	5	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	29	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	30	1	Submitted	submit	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	30	2	12/2012	12/2012	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	30	3	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	30	4	Revised	Revised	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	30	5	3/2013	3/2013	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	30	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	30	7	5/2013	5/2013	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	30	8	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	30	9	Published	Published	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	30	10	7/2013	7/2013	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	30	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	31	1	c	c	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	31	2	©	©	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	31	3	2013	2013	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	31	4	Association	Association	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	31	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	31	6	Computational	Computational	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	31	7	Linguistics	Linguistics	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	31	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	32	1	zani	zani	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	2	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	3	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	4	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	32	6	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	32	7	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	32	8	Mooney	Mooney	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	32	10	Roy	Roy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	32	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	32	12	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	32	13	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	32	14	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	32	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	32	16	c	c	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	32	17	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	32	18	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	32	19	easy	easy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	32	20	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	32	21	read	read	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	32	22	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	32	23	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	32	24	target	target	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	32	25	readership	readership	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	32	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	33	1	Shorter	short	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	33	2	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	33	3	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	4	Flesch	Flesch	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	33	6	1948	1948	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	33	7	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	33	9	less	less	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	33	10	complex	complex	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	33	11	syntax	syntax	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	33	12	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	13	Schwarm	Schwarm	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	33	15	Ostendorf	Ostendorf	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	33	17	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	33	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	33	20	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	33	21	cohesion	cohesion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	33	22	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	33	23	sen-	sen-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	33	24	tences	tence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	33	25	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	26	Graesser	Graesser	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	27	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	28	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	29	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	33	31	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	33	32	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	33	33	typically	typically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	33	34	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	33	35	eas-	eas-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	33	36	ier	i	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	33	37	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	33	38	more	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	33	39	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	33	40	readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	33	41	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	33	42	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	33	43	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	34	1	Less	less	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	34	2	understood	understand	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	34	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	34	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	34	5	question	question	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	34	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	34	7	what	what	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	34	8	makes	make	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	34	9	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	34	10	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	34	11	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	34	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	34	13	beautifully	beautifully	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	34	14	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	34	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	35	1	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	35	2	early	early	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	35	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	35	4	influential	influential	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	35	5	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	6	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	7	readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	35	9	Flesch	Flesch	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	35	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	35	11	1948	1948	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	35	12	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	35	13	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	35	14	computed	compute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	35	15	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	35	16	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	17	measure	measure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	18	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	35	20	hypoth-	hypoth-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	35	21	esis	esis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	22	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	35	23	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	35	24	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	35	25	would	would	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	35	26	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	35	27	easier	easy	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	35	28	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	35	29	read	read	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	35	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	36	1	More	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	36	2	recently	recently	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	36	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	36	4	McIntyre	McIntyre	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	36	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	36	6	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	36	7	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	36	8	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	36	9	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	36	10	found	find	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	36	11	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	12	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	13	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	14	ratings	rating	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	16	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	18	fairy	fairy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	19	tales	tale	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	20	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	36	21	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	36	22	successfully	successfully	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	36	23	predicted	predict	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	24	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	36	25	token	token	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	36	26	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	36	27	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	28	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	29	re-	re-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	36	30	lated	late	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	36	31	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	32	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	36	33	items	item	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	34	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	36	35	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	36	36	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	37	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	36	38	psy-	psy-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	39	cholinguistic	cholinguistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	36	40	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	36	41	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	37	1	But	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	37	2	large	large	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	37	3	scale	scale	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	4	studies	study	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	37	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	6	in-	in-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	37	7	terest	ter	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	37	8	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	37	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	10	adult	adult	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	11	educated	educated	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	37	12	readers	reader	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	37	13	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	37	14	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	37	15	been	be	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	16	carried	carry	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	37	17	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	37	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	38	1	Further	further	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	38	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	38	3	there	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	38	4	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	38	5	been	be	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	6	little	little	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	38	7	attempts	attempt	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	38	8	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	38	9	measure	measure	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	38	10	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	11	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	38	14	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	15	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	38	16	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	38	17	setting	setting	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	38	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	39	1	But	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	39	2	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	4	reasonable	reasonable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	39	6	expect	expect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	39	7	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	8	properties	property	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	39	9	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	10	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	39	12	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	13	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	39	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	15	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	39	16	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	17	should	should	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	39	18	contribute	contribute	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	39	19	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	39	21	prediction	prediction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	23	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	24	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	25	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	39	26	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	27	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	28	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	39	29	domain-	domain-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	30	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	31	spelling	spelling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	32	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	39	33	grammar	grammar	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	34	correction	correction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	35	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	39	36	Cucerzan	Cucerzan	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	37	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	39	38	Brill	Brill	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	39	40	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	39	41	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	39	42	Bao	Bao	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	43	et	et	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	39	44	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	45	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	46	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	39	47	2011	2011	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	39	48	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	39	49	Dale	Dale	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	50	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	39	51	Kilgar-	Kilgar-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	52	riff	riff	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	53	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	39	54	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	39	55	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	39	56	techniques	technique	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	39	57	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	39	58	been	be	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	39	59	successful	successful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	39	60	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	40	1	Here	here	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	40	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	40	3	address	address	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	40	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	40	5	above	above	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	40	6	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	40	7	issues	issue	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	40	8	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	9	develop-	develop-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	40	10	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	40	11	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	40	12	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	13	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	14	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	40	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	40	16	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	40	17	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	40	18	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	19	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	20	specifically	specifically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	40	21	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	22	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	23	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	40	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	41	1	Central	central	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	41	2	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	41	4	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	5	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	41	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	41	7	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	9	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	10	news	news	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	11	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	41	12	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	13	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	41	14	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	41	15	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	41	16	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	17	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	18	popular	popular	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	41	19	journalists	journalist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	41	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	41	21	typical	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	41	22	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	41	23	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	24	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	25	columns	column	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	41	26	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	41	27	Section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	41	28	2	2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	41	29	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	41	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	42	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	2	introduce	introduce	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	42	4	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	6	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	42	8	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	9	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	42	10	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	11	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	12	beautiful	beautiful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	13	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	42	15	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	16	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	42	18	affective	affective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	19	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	42	21	Section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	22	3	3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	42	23	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	42	24	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	42	25	show	show	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	26	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	27	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	28	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	29	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	30	predictive	predictive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	42	31	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	42	32	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	42	33	20	20	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	42	34	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	35	above	above	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	36	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	42	37	baseline	baseline	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	38	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	42	39	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	40	distinguish-	distinguish-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	42	41	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	42	42	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	42	43	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	44	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	42	45	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	42	46	Section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	42	47	4	4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	42	48	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	42	49	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	43	1	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	43	2	final	final	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	43	3	sys-	sys-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	4	tem	tem	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	5	combines	combine	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	43	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	43	7	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	43	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	9	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	43	11	genre-	genre-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	12	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	43	13	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	43	14	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	15	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	43	16	proposed	propose	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	18	identifying	identify	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	43	19	readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	43	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	43	21	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	43	22	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	43	23	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	24	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	43	25	topically	topically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	43	26	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	43	27	arti-	arti-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	28	cles	cles	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	43	29	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	43	30	giving	give	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	43	31	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	43	32	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	33	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	34	84	84	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	43	35	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	36	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	43	37	Section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	43	38	5	5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	43	39	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	43	40	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	44	1	2	2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	44	2	Article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	3	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	4	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	5	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	44	6	corpus1	corpus1	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	7	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	44	8	chosen	choose	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	9	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	44	10	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	44	12	larger	large	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	44	13	New	New	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	14	York	York	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	15	Times	Times	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	44	17	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	44	19	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	44	21	Sandhaus	Sandhaus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	44	23	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	44	24	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	44	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	44	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	44	27	latter	latter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	28	containing	contain	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	44	29	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	44	30	wealth	wealth	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	31	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	32	metadata	metadata	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	33	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	34	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	44	35	1Available	1available	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	44	36	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	37	http://www.cis.upenn.edu/	http://www.cis.upenn.edu/	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	38	˜nlp	˜nlp	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	39	/	/	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	44	40	corpora	corpora	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	41	/	/	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	44	42	scinewscorpus.html	scinewscorpus.html	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	44	43	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	44	including	include	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	44	45	author	author	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	46	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	47	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	44	48	manually	manually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	44	49	assigned	assign	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	50	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	44	51	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	44	52	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	45	1	2.1	2.1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	45	2	General	General	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	45	3	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	4	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	45	5	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	45	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	45	8	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	45	9	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	45	10	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	11	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	45	12	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	45	13	contributions	contribution	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	45	14	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	45	16	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	45	17	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	18	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	45	19	whose	whose	WP$
work_hea3gijzhnc2rl4sxgyj6pvghm	45	20	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	21	appeared	appear	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	45	22	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	23	“	"	``
work_hea3gijzhnc2rl4sxgyj6pvghm	45	24	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	45	25	Best	good	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	45	26	American	american	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	45	27	Science	Science	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	45	28	Writing	Writing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	45	29	”	"	''
work_hea3gijzhnc2rl4sxgyj6pvghm	45	30	anthology	anthology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	31	published	publish	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	32	annually	annually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	45	33	since	since	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	45	34	1999	1999	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	45	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	46	1	Articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	46	2	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	46	4	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	5	columns	column	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	46	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	7	leading	leading	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	46	8	newspapers	newspaper	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	46	9	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	46	10	nominated	nominate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	46	12	expert	expert	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	46	13	journalists	journalist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	46	14	choose	choose	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	46	15	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	46	16	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	17	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	46	18	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	46	19	exceptional	exceptional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	46	20	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	46	21	appear	appear	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	46	22	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	46	23	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	46	24	anthologies	anthology	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	46	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	47	1	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	47	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	3	63	63	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	47	4	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	5	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	47	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	8	anthology	anthology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	9	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	47	10	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	11	years	year	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	47	12	1999	1999	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	47	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	47	14	2007	2007	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	47	15	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	47	16	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	17	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	18	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	47	19	part	part	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	20	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	22	digital	digital	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	47	23	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	24	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	25	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	47	26	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	27	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	47	28	form	form	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	47	29	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	30	seed	seed	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	31	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	32	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	33	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	47	34	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	47	35	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	47	36	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	47	37	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	48	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	48	2	further	further	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	48	3	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	48	4	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	48	6	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	48	7	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	48	8	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	9	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	48	10	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	48	11	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	48	13	contributed	contribute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	48	14	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	15	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	48	16	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	48	18	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	48	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	48	21	seed	seed	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	48	22	examples	example	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	48	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	49	1	Science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	2	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	49	3	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	4	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	49	5	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	49	6	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	49	7	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	8	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	49	9	seed	seed	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	10	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	11	form	form	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	49	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	49	13	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	49	14	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	49	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	50	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	50	2	perform	perform	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	50	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	50	4	expansion	expansion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	50	5	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	50	6	first	first	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	50	7	creat-	creat-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	50	8	ing	ing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	50	9	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	50	10	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	50	11	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	50	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	50	13	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	50	14	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	50	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	51	1	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	51	2	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	51	3	no	no	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	51	4	single	single	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	51	5	meta	meta	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	51	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	51	7	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	51	8	tag	tag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	51	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	51	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	51	11	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	51	12	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	51	13	refers	refer	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	51	14	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	51	15	all	all	PDT
work_hea3gijzhnc2rl4sxgyj6pvghm	51	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	51	17	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	51	18	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	51	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	52	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	52	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	52	3	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	52	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	52	5	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	6	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	52	7	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	52	9	seed	seed	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	52	11	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	12	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	52	13	initial	initial	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	52	14	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	16	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	52	17	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	52	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	53	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	53	2	then	then	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	53	3	compute	compute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	53	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	53	5	minimal	minimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	53	6	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	53	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	53	8	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	53	9	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	53	10	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	53	11	cover	cover	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	53	12	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	53	13	best	good	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	53	14	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	53	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	54	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	54	2	greedily	greedily	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	54	3	add	add	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	54	4	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	54	5	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	7	minimal	minimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	54	8	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	54	10	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	11	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	12	iteration	iteration	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	13	choosing	choose	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	54	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	15	tag	tag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	16	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	17	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	54	18	present	present	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	54	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	21	majority	majority	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	54	23	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	54	24	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	54	25	re-	re-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	54	26	main	main	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	54	27	uncovered	uncover	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	54	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	55	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	55	2	minimal	minimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	55	3	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	55	4	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	55	5	14	14	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	55	6	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	55	7	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	55	8	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	55	9	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	55	10	Medicine	Medicine	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	55	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	55	12	Health	Health	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	55	13	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	55	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	55	15	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	55	16	Space	space	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	55	17	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	55	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	55	19	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	55	20	Research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	55	21	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	55	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	55	23	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	55	24	Physics	Physics	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	55	25	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	55	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	55	27	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	55	28	Evolution	Evolution	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	55	29	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	55	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	56	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	56	2	collect	collect	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	56	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	56	4	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	56	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	56	6	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	56	7	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	56	8	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	56	9	at	at	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	56	10	least	least	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	56	11	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	56	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	56	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	56	14	minimal	minimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	56	15	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	56	16	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	56	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	57	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	57	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	57	3	even	even	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	57	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	57	5	cursory	cursory	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	57	6	mention	mention	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	57	9	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	10	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	11	results	result	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	57	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	57	14	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	15	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	57	16	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	17	tag	tag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	18	being	be	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	57	19	assigned	assign	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	20	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	57	22	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	57	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	58	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	58	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	58	3	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	58	4	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	58	5	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	58	6	dictionary	dictionary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	8	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	9	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	58	10	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	11	terms	term	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	58	12	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	58	13	determine	determine	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	58	14	whether	whether	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	58	16	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	17	passes	pass	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	58	18	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	58	19	minimum	minimum	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	58	20	threshold	threshold	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	21	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	22	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	23	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	58	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	59	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	59	2	created	create	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	59	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	59	4	dic-	dic-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	59	5	tionary	tionary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	59	6	manually	manually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	59	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	59	8	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	59	9	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	59	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	59	11	following	follow	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	59	12	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	59	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	59	14	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	59	15	morphological	morphological	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	59	16	variants	variant	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	59	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	59	18	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	59	19	63	63	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	59	20	items	item	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	59	21	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	59	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	60	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	60	2	used	use	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	60	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	60	4	intuition	intuition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	60	5	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	60	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	60	7	few	few	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	60	8	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	60	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	60	10	re-	re-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	60	11	search	search	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	60	12	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	60	13	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	60	14	create	create	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	60	15	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	60	16	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	60	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	61	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	61	2	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	61	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	61	4	shown	show	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	61	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	61	6	capital	capital	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	61	7	letters	letter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	61	8	below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	61	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	62	1	PEOPLE	PEOPLE	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	62	2	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	3	researcher	researcher	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	5	scientist	scientist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	7	physicist	physicist	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	9	biologist	biologist	UH
work_hea3gijzhnc2rl4sxgyj6pvghm	62	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	11	economist	economist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	13	anthropologist	anthropologist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	15	environmentalist	environmentalist	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	17	linguist	linguist	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	19	professor	professor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	21	dr	dr	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	23	student	student	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	24	PROCESS	PROCESS	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	25	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	26	discover	discover	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	62	27	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	28	found	find	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	29	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	30	experiment	experiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	32	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	33	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	34	finding	finding	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	35	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	36	study	study	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	38	question	question	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	40	project	project	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	41	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	42	discuss	discuss	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	62	43	TOPIC	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	44	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	45	biology	biology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	46	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	47	physics	physics	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	48	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	49	chemistry	chemistry	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	50	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	51	anthropology	anthropology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	52	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	53	primatology	primatology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	54	342	342	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	62	55	PUBLICATIONS	publications	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	56	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	57	report	report	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	58	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	59	published	publish	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	60	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	61	journal	journal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	62	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	63	paper	paper	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	64	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	65	author	author	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	66	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	67	issue	issue	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	68	OTHER	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	69	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	70	human	human	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	71	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	72	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	73	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	74	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	75	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	76	knowledge	knowledge	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	77	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	78	university	university	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	79	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	80	lab-	lab-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	81	oratory	oratory	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	82	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	83	lab	lab	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	84	ENDINGS	ending	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	62	85	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	62	86	-ology	-ology	NFP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	87	-gist	-gist	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	88	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	89	-list	-list	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	62	90	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	91	-mist	-mist	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	92	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	93	-uist	-uist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	94	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	62	95	-phy	-phy	.
work_hea3gijzhnc2rl4sxgyj6pvghm	62	96	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	62	97	items	item	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	62	98	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	99	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	62	100	ENDINGS	ENDINGS	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	101	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	102	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	62	103	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	104	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	62	105	match	match	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	62	106	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	62	107	suffixes	suffix	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	62	108	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	63	1	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	63	2	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	63	4	considered	consider	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	5	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	63	7	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	8	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	9	at	at	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	63	10	least	least	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	63	11	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	63	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	13	its	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	63	14	tokens	token	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	63	15	match	match	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	63	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	63	17	dictionary	dictionary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	63	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	20	addition	addition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	63	22	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	23	least	least	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	63	24	5	5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	63	25	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	63	26	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	63	27	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	63	29	dictionary	dictionary	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	63	30	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	63	31	matched	match	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	63	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	64	1	Since	since	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	64	3	time	time	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	4	span	span	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	64	7	best	good	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	64	8	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	64	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	64	10	1999	1999	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	64	11	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	12	2007	2007	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	64	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	64	14	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	64	15	limit	limit	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	64	16	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	64	17	col-	col-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	64	18	lection	lection	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	19	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	20	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	64	21	timespan	timespan	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	64	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	65	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	65	2	addition	addition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	65	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	65	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	65	5	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	65	6	con-	con-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	65	7	sider	sider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	65	8	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	65	9	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	65	10	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	65	11	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	65	12	least	least	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	65	13	500	500	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	65	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	65	15	long	long	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	65	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	66	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	66	2	filtered	filter	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	3	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	5	23,709	23,709	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	66	6	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	66	7	form	form	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	66	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	66	9	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	66	10	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	12	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	13	journalism	journalism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	66	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	67	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	67	2	63	63	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	67	3	seed	seed	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	4	samples	sample	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	67	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	6	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	67	7	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	8	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	67	9	con-	con-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	10	tributed	tribute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	11	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	67	12	about	about	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	67	13	40	40	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	67	14	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	67	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	68	1	Some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	68	2	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	3	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	68	4	multiple	multiple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	68	5	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	6	selected	select	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	68	9	best	good	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	10	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	11	book	book	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	12	series	series	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	68	14	supporting	support	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	68	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	68	16	idea	idea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	17	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	18	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	68	19	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	20	produce	produce	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	68	21	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	68	22	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	23	pieces	piece	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	24	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	68	25	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	68	26	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	68	27	considered	consider	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	28	distinct	distinct	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	68	29	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	68	30	typical	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	68	31	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	68	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	69	1	Separating	separate	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	69	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	69	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	69	4	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	69	5	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	69	6	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	69	7	gives	give	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	69	8	us	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	69	9	3,467	3,467	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	69	10	extra	extra	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	69	11	samples	sample	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	69	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	69	13	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	69	14	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	69	15	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	69	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	70	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	70	2	total	total	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	70	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	70	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	70	5	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	70	6	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	70	7	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	70	8	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	70	9	3,530	3,530	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	70	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	70	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	71	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	71	2	remaining	remain	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	71	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	71	4	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	71	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	71	6	rele-	rele-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	71	7	vant	vant	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	71	8	set	set	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	71	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	71	10	20,242	20,242	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	71	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	71	12	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	71	13	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	71	14	about	about	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	71	15	3000	3000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	71	16	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	71	17	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	71	18	form	form	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	71	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	71	20	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	71	21	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	71	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	72	1	2.2	2.2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	72	2	Topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	72	4	paired	pair	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	5	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	72	6	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	72	7	general	general	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	72	8	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	10	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	11	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	12	introduced	introduce	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	13	so	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	72	14	far	far	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	72	15	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	72	16	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	72	17	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	18	diverse	diverse	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	72	19	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	72	20	including	include	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	72	21	bi-	bi-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	72	22	ology	ology	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	23	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	72	24	astronomy	astronomy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	72	26	religion	religion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	72	27	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	72	28	sports	sport	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	72	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	73	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	73	2	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	73	3	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	73	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	73	5	TYPICAL	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	73	6	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	73	7	created	create	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	8	above	above	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	9	al-	al-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	73	10	low	low	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	11	us	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	73	12	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	73	13	study	study	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	73	14	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	15	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	16	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	17	regard	regard	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	18	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	19	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	73	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	74	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	74	2	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	74	3	typical	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	74	4	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	5	retrieval	retrieval	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	6	sce-	sce-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	74	7	nario	nario	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	74	8	would	would	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	74	9	involve	involve	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	74	10	comparison	comparison	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	11	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	74	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	74	15	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	74	16	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	74	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	74	18	i.e.	i.e.	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	75	1	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	75	2	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	75	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	75	4	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	75	5	query	query	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	75	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	76	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	76	2	investigate	investigate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	76	3	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	76	4	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	5	differentiation	differentiation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	6	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	76	7	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	76	8	done	do	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	9	within	within	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	10	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	76	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	76	12	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	76	13	created	create	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	76	14	another	another	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	76	15	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	16	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	76	17	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	76	18	paired	pair	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	76	19	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	76	20	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	21	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	76	22	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	76	23	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	76	24	TYPICAL	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	76	25	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	76	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	77	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	2	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	77	3	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	4	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	77	6	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	77	7	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	77	8	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	77	10	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	77	11	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	77	12	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	13	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	14	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	77	15	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	77	16	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	77	18	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	77	19	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	77	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	78	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	78	2	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	78	4	computed	compute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	5	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	6	comparing	compare	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	78	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	78	8	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	78	10	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	78	11	computed	compute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	12	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	78	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	78	14	loglikelihood	loglikelihood	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	15	ratio	ratio	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	16	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	78	18	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	78	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	78	20	Hovy	Hovy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	78	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	78	22	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	78	23	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	78	24	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	78	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	78	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	78	27	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	78	28	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	78	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	79	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	79	2	retain	retain	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	79	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	79	4	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	79	5	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	79	6	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	79	7	TYPICAL	typical	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	79	8	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	79	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	79	10	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	79	11	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	79	12	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	79	13	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	79	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	80	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	80	2	enumerate	enumerate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	80	3	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	80	4	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	80	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	6	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	80	7	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	80	8	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	9	matched	match	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	10	up	up	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	80	11	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	80	12	ARTICLES	ARTICLES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	80	13	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	80	14	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	80	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	16	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	17	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	80	18	giving	give	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	80	19	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	80	20	total	total	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	80	22	35,300	35,300	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	80	23	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	80	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	81	1	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	81	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	81	3	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	81	4	distinguishing	distinguish	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	81	5	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	81	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	81	7	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	81	8	cor-	cor-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	81	9	pus	pus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	81	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	82	1	First	first	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	82	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	82	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	82	4	average	average	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	82	5	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	82	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	82	7	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	82	8	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	82	9	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	82	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	83	1	They	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	83	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	83	3	unlikely	unlikely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	83	4	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	83	5	have	have	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	83	6	spelling	spelling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	83	8	grammar	grammar	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	83	10	ba-	ba-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	83	11	sic	sic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	12	organization	organization	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	13	problems	problem	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	83	14	allowing	allow	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	83	15	us	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	83	16	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	83	17	investigate	investigate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	83	18	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	19	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	20	rather	rather	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	83	21	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	83	23	detection	detection	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	24	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	83	25	errors	error	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	83	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	84	1	Second	second	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	84	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	84	4	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	5	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	6	more	more	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	84	7	realistic	realistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	8	samples	sample	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	84	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	10	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	11	differences	difference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	84	12	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	13	IR	IR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	84	14	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	84	15	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	16	recommen-	recommen-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	17	dation	dation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	18	compared	compare	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	19	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	20	prior	prior	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	21	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	84	23	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	84	24	system	system	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	25	pro-	pro-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	26	duced	duce	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	84	27	texts	text	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	84	28	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	84	29	permuted	permute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	84	30	versions	version	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	84	31	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	32	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	84	33	original	original	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	84	34	ar-	ar-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	35	ticle	ticle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	36	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	84	37	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	38	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	39	proxies	proxy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	84	40	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	41	lower	low	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	84	42	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	43	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	84	44	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	85	1	2.3	2.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	85	2	Tasks	task	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	85	3	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	85	4	perform	perform	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	85	5	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	85	6	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	85	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	85	8	classification	classification	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	85	9	tasks	task	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	85	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	86	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	86	2	divide	divide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	86	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	86	4	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	5	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	6	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	86	8	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	9	sets	set	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	86	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	11	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	86	12	tasks	task	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	86	13	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	86	15	following	following	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	86	16	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	86	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	87	1	Any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	87	2	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	3	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	87	4	Here	here	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	87	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	87	6	goal	goal	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	7	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	87	8	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	87	9	separate	separate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	87	10	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	87	11	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	87	12	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	87	13	versus	versus	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	14	TYPICAL	typical	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	15	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	87	16	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	17	regard	regard	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	18	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	19	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	87	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	88	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	88	2	setting	set	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	88	3	roughly	roughly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	88	4	corresponds	correspond	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	88	5	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	6	picking	pick	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	88	7	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	88	8	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	88	9	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	88	10	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	11	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	12	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	88	13	archive	archive	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	14	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	88	15	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	88	16	day	day	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	17	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	88	18	newspaper	newspaper	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	88	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	89	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	89	2	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	89	3	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	89	4	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	89	5	3,430	3,430	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	89	6	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	89	7	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	89	8	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	89	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	89	10	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	89	11	randomly	randomly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	89	12	sample	sample	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	89	13	3,430	3,430	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	89	14	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	89	15	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	89	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	89	17	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	89	18	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	89	19	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	89	20	comprise	comprise	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	89	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	89	22	negative	negative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	89	23	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	89	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	90	1	Same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	90	2	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	90	3	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	90	4	Here	here	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	90	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	90	6	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	90	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	90	8	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	90	9	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	90	10	paired	pair	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	90	11	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	90	12	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	90	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	90	14	TYPICAL	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	90	15	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	90	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	91	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	91	2	goal	goal	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	91	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	91	4	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	91	5	predict	predict	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	91	6	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	91	7	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	91	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	91	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	91	10	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	91	11	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	91	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	91	13	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	91	14	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	91	15	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	91	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	92	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	92	2	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	92	4	closer	close	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	92	5	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	92	7	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	8	retrieval	retrieval	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	9	setting	setting	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	92	11	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	92	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	92	13	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	92	14	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	15	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	92	17	retrieved	retrieve	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	18	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	19	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	92	20	user	user	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	21	query	query	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	92	23	need	need	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	92	24	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	92	25	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	92	26	distinguished	distinguish	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	27	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	28	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	92	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	93	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	93	2	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	93	3	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	93	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	93	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	93	6	selected	select	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	93	7	34,300	34,300	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	93	8	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	93	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	94	1	Development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	94	2	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	94	3	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	94	4	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	94	5	randomly	randomly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	94	6	selected	select	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	94	7	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	94	8	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	94	9	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	94	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	94	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	94	12	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	94	13	paired	pair	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	94	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	94	15	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	94	16	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	94	17	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	94	18	TYPICAL	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	94	19	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	94	20	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	94	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	94	22	topic	topic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	94	23	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	94	24	normalized	normalized	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	94	25	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	94	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	95	1	Overall	overall	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	95	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	95	3	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	95	4	constitute	constitute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	95	5	1,000	1,000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	95	6	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	95	7	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	95	8	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	95	9	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	95	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	95	11	developing	develop	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	95	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	95	13	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	95	14	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	95	15	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	95	16	classifier	classifier	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	95	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	96	1	From	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	96	2	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	96	3	selected	select	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	96	4	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	96	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	96	6	take	take	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	96	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	96	8	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	96	9	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	96	10	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	96	11	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	96	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	96	13	sample	sample	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	96	14	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	96	15	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	96	16	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	96	17	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	96	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	96	19	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	96	20	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	96	21	making	make	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	96	22	up	up	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	96	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	96	24	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	96	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	97	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	97	2	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	97	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	97	4	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	97	5	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	97	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	97	7	tune	tune	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	97	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	97	9	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	97	10	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	97	11	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	97	12	classifier	classifier	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	97	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	98	1	3	3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	98	2	Facets	facet	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	98	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	4	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	5	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	6	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	7	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	98	8	section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	98	10	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	98	11	discuss	discuss	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	98	12	six	six	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	98	13	prominent	prominent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	98	14	facets	facet	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	98	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	16	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	17	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	18	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	98	19	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	98	20	hypothesized	hypothesize	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	98	21	will	will	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	98	22	have	have	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	98	23	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	98	24	impact	impact	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	25	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	26	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	27	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	98	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	99	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	99	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	99	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	99	4	presence	presence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	6	passages	passage	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	99	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	8	highly	highly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	99	9	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	99	10	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	99	12	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	99	13	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	99	14	oriented	orient	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	15	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	99	17	use	use	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	18	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	19	beautiful	beautiful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	99	20	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	99	22	sub	sub	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	23	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	99	24	genres	genre	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	99	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	99	26	sen-	sen-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	99	27	timent	timent	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	28	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	99	29	affect	affect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	99	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	99	31	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	99	32	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	99	33	depth	depth	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	34	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	35	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	36	descrip-	descrip-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	99	37	tion	tion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	99	38	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	100	1	Several	several	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	2	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	3	properties	property	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	100	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	5	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	6	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	7	could	could	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	100	8	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	100	9	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	100	10	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	11	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	12	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	13	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	14	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	100	16	use	use	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	18	343	343	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	100	19	humor	humor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	100	21	metaphor	metaphor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	100	23	suspense	suspense	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	24	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	100	25	clarity	clarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	26	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	27	explana-	explana-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	28	tions	tion	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	100	29	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	100	30	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	100	31	plan	plan	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	100	32	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	100	33	explore	explore	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	100	34	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	100	35	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	36	future	future	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	100	37	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	100	38	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	101	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	101	2	describe	describe	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	101	3	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	101	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	101	5	computed	compute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	101	6	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	101	7	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	101	8	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	101	9	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	101	10	property	property	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	101	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	101	12	tested	test	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	101	13	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	101	14	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	101	15	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	101	16	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	101	17	dis-	dis-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	101	18	tributed	tribute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	101	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	101	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	101	21	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	101	22	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	101	23	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	101	24	TYPICAL	typical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	101	25	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	101	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	102	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	102	2	do	do	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	102	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	102	4	analysis	analysis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	102	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	102	6	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	102	7	randomly	randomly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	102	8	sampled	sample	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	102	9	1,000	1,000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	102	10	ar-	ar-	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	102	11	ticles	ticle	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	102	12	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	102	13	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	102	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	102	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	102	16	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	102	17	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	102	18	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	102	19	representa-	representa-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	102	20	tive	tive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	102	21	examples	example	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	102	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	103	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	103	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	103	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	4	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	6	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	7	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	9	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	103	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	103	12	use	use	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	103	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	14	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	103	15	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	103	16	sided	sided	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	103	17	t	t	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	18	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	103	19	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	20	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	103	21	check	check	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	103	22	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	24	mean	mean	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	103	25	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	26	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	103	28	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	29	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	103	30	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	103	31	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	32	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	103	33	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	34	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	103	35	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	103	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	104	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	104	2	p	p	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	104	4	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	5	less	less	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	104	6	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	7	0.05	0.05	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	104	8	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	104	9	taken	take	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	10	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	11	in-	in-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	104	12	dicate	dicate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	104	13	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	104	14	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	104	15	trend	trend	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	16	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	104	18	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	104	21	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	104	22	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	104	23	versus	versus	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	104	24	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	104	25	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	104	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	105	1	Note	note	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	105	2	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	105	4	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	5	computation	computation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	6	step	step	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	7	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	105	8	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	105	9	tuned	tune	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	105	12	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	13	prediction	prediction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	14	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	16	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	105	17	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	105	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	106	1	Rather	rather	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	106	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	106	3	aim	aim	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	106	4	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	106	5	represent	represent	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	106	6	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	106	7	facet	facet	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	106	8	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	106	9	accurately	accurately	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	106	10	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	106	11	possible	possible	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	106	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	107	1	Ideally	ideally	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	107	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	107	3	would	would	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	107	4	require	require	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	107	5	manual	manual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	107	6	anno-	anno-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	107	7	tations	tation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	107	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	107	9	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	107	10	facet	facet	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	107	11	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	107	12	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	107	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	107	14	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	107	15	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	107	16	etc	etc	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	107	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	107	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	108	1	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	108	2	achieve	achieve	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	108	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	108	4	goal	goal	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	108	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	109	1	At	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	2	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	109	3	time	time	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	109	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	109	6	simply	simply	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	109	7	check	check	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	109	8	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	109	9	chosen	choose	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	10	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	109	11	’	’	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	109	12	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	109	13	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	14	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	109	15	random	random	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	109	16	collection	collection	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	18	snippets	snippet	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	109	19	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	20	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	109	21	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	22	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	109	23	check	check	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	109	24	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	25	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	109	26	behave	behave	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	109	27	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	28	intended	intend	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	29	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	30	resorting	resort	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	109	31	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	109	32	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	109	33	annotations	annotation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	109	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	110	1	3.1	3.1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	110	2	Visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	110	3	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	5	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	110	6	Some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	110	7	texts	text	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	110	8	create	create	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	110	9	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	110	10	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	110	13	reader	reader	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	14	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	110	15	mind	mind	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	110	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	111	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	111	2	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	111	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	111	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	111	5	snippet	snippet	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	111	6	below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	111	7	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	111	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	111	9	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	111	10	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	111	11	effect	effect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	111	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	112	1	When	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	112	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	112	3	sea	sea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	4	lions	lion	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	112	5	approached	approach	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	112	6	close	close	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	112	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	112	8	seemingly	seemingly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	112	9	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	112	10	curious	curious	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	112	11	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	12	us	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	112	13	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	14	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	112	15	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	112	16	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	17	them	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	112	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	112	19	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	112	20	big	big	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	112	21	brown	brown	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	112	22	eyes	eye	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	112	23	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	112	24	encircled	encircle	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	25	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	26	light	light	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	112	27	fur	fur	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	28	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	112	29	looked	look	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	112	30	like	like	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	31	makeup	makeup	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	112	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	113	1	One	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	113	2	sea	sea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	3	lion	lion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	4	played	play	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	113	5	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	113	7	conch	conch	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	8	shell	shell	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	9	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	10	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	11	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	113	12	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	113	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	113	14	ball	ball	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	113	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	114	1	Such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	114	2	vivid	vivid	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	114	3	descriptions	description	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	114	4	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	114	5	engage	engage	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	114	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	114	7	entertain	entertain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	114	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	114	9	reader	reader	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	114	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	115	1	Kosslyn	Kosslyn	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	115	2	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	115	3	1980	1980	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	115	4	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	115	5	found	find	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	115	6	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	115	7	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	115	8	spon-	spon-	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	115	9	taneously	taneously	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	115	10	form	form	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	115	11	images	image	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	115	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	115	13	concrete	concrete	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	115	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	115	15	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	115	16	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	115	17	hear	hear	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	115	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	115	19	use	use	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	115	20	them	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	115	21	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	115	22	answer	answer	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	115	23	questions	question	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	115	24	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	115	25	perform	perform	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	115	26	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	115	27	tasks	task	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	115	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	116	1	Books	book	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	116	2	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	4	student	student	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	5	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	6	jour-	jour-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	7	nalists	nalist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	116	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	116	9	Blum	Blum	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	116	10	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	116	11	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	116	12	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	116	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	116	14	2006	2006	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	116	15	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	116	16	Stocking	stocking	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	116	18	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	116	19	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	116	20	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	116	21	em-	em-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	116	22	phasize	phasize	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	116	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	116	24	importance	importance	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	116	26	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	116	27	descriptions	description	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	116	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	117	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	117	2	measure	measure	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	117	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	117	4	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	117	5	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	117	8	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	9	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	10	count-	count-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	117	11	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	117	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	117	13	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	117	15	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	117	16	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	117	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	118	1	Currently	currently	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	118	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	118	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	118	4	only	only	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	118	5	resource	resource	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	118	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	118	7	imagery	imagery	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	118	8	ratings	rating	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	118	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	118	10	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	118	11	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	118	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	118	13	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	118	14	psycholinguistic	psycholinguistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	118	15	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	118	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	118	17	Wilson	Wilson	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	118	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	118	19	1988	1988	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	118	20	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	118	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	119	1	It	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	119	2	con-	con-	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	119	3	tains	tain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	119	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	5	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	7	3,394	3,394	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	119	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	119	9	rated	rate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	11	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	119	12	ability	ability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	13	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	119	14	invoke	invoke	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	15	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	16	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	119	18	so	so	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	20	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	21	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	119	22	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	23	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	119	24	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	25	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	119	26	highly	highly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	27	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	119	28	along	along	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	29	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	119	30	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	119	31	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	119	32	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	119	33	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	34	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	119	35	at	at	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	36	all	all	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	119	37	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	120	1	With	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	2	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	120	3	cutoff	cutoff	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	4	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	6	adopted	adopt	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	120	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	9	4.5	4.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	120	12	Gilhooly	Gilhooly	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	13	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	120	14	Logie	Logie	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	120	16	350	350	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	120	19	Bristol	Bristol	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	20	Norms2	Norms2	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	21	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	22	2The	2the	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	23	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	120	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	120	25	resource	resource	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	26	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	120	27	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	120	28	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	120	29	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	30	lists	list	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	120	31	—	—	:
work_hea3gijzhnc2rl4sxgyj6pvghm	120	32	obtain	obtain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	120	33	1,966	1,966	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	120	34	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	120	35	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	120	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	121	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	121	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	121	3	coverage	coverage	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	121	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	121	5	that	that	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	121	6	lexicon	lexicon	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	121	7	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	121	8	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	121	9	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	121	10	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	121	11	low	low	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	121	12	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	121	13	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	121	14	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	121	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	122	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	2	collect	collect	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	122	4	larger	large	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	122	5	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	7	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	122	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	122	9	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	10	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	122	11	cor-	cor-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	12	pus	pus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	14	tagged	tag	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	15	images	image	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	122	16	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	122	18	ESP	ESP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	19	game	game	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	122	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	122	21	von	von	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	22	Ahn	Ahn	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	23	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	122	24	Dabbish	Dabbish	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	122	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	122	26	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	122	27	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	122	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	123	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	123	2	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	123	3	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	123	4	83,904	83,904	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	123	5	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	123	6	images	image	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	123	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	123	8	27,466	27,466	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	123	9	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	123	10	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	123	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	124	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	124	2	average	average	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	124	3	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	124	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	124	5	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	124	6	per	per	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	124	7	picture	picture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	124	8	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	124	9	14.5	14.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	124	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	125	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	125	2	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	125	3	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	125	4	collected	collect	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	125	7	game	game	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	8	setting	set	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	125	9	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	125	10	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	125	11	users	user	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	125	12	individ-	individ-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	125	13	ually	ually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	125	14	saw	see	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	125	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	125	16	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	125	17	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	125	19	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	125	20	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	125	21	guess	guess	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	125	22	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	125	23	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	24	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	125	25	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	125	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	126	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	126	2	players	player	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	126	3	increased	increase	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	126	4	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	126	5	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	126	6	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	126	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	126	8	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	126	9	guessed	guess	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	126	10	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	126	11	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	126	12	player	player	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	126	13	matched	match	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	126	14	that	that	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	126	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	126	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	126	17	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	126	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	127	1	Due	due	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	2	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	127	4	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	5	annotation	annotation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	6	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	127	8	there	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	127	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	10	considerable	considerable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	11	noise	noise	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	127	13	non	non	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	14	-	-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	15	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	127	16	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	127	17	assigned	assign	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	18	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	127	19	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	127	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	128	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	128	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	128	3	performed	perform	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	128	4	filtering	filter	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	128	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	128	6	find	find	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	128	7	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	128	8	pre-	pre-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	128	9	cision	cision	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	128	10	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	128	11	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	128	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	128	13	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	128	14	group	group	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	128	15	them	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	128	16	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	128	17	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	128	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	129	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	3	Latent	Latent	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	4	Dirichlet	Dirichlet	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	5	Allocation	Allocation	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	6	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	129	7	Blei	Blei	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	8	et	et	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	129	9	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	10	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	129	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	129	12	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	129	13	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	129	14	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	129	15	cluster	cluster	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	129	16	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	129	17	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	129	18	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	129	19	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	129	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	130	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	130	2	treat	treat	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	130	3	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	130	4	picture	picture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	5	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	130	7	document	document	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	130	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	130	10	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	130	11	assigned	assign	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	12	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	130	14	picture	picture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	15	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	130	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	130	17	document	document	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	130	18	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	130	19	contents	content	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	130	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	131	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	131	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	131	3	sym-	sym-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	131	4	metric	metric	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	131	5	priors	prior	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	131	6	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	7	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	8	0.01	0.01	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	131	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	10	both	both	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	131	11	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	12	mixture	mixture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	131	14	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	15	distribution	distribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	16	within	within	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	17	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	131	18	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	131	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	132	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	132	2	filter	filter	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	132	3	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	132	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	132	5	30	30	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	132	6	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	132	7	common	common	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	132	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	132	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	132	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	132	11	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	132	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	132	13	words	word	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	132	14	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	132	15	ap-	ap-	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	132	16	pear	pear	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	132	17	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	132	18	less	less	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	132	19	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	132	20	four	four	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	132	21	pictures	picture	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	132	22	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	132	23	images	image	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	132	24	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	132	25	fewer	few	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	132	26	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	132	27	five	five	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	132	28	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	132	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	133	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	133	2	remaining	remain	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	133	3	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	133	4	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	5	clustered	cluster	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	133	6	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	133	7	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	133	8	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	133	9	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	133	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	133	11	Stanford	Stanford	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	12	Topic	Topic	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	13	Modeling	Modeling	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	14	Toolbox3	Toolbox3	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	15	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	133	16	Ramage	Ramage	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	17	et	et	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	133	18	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	19	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	133	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	133	21	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	133	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	133	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	134	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	134	2	did	do	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	134	3	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	134	4	tune	tune	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	134	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	134	6	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	8	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	134	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	134	10	choose	choose	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	134	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	134	12	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	14	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	134	15	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	16	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	134	18	intuition	intuition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	19	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	134	21	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	134	23	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	134	24	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	134	25	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	134	26	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	134	27	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	134	28	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	134	29	small	small	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	134	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	135	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	135	2	select	select	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	135	3	clean	clean	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	4	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	5	clusters	cluster	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	135	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	135	7	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	135	8	make	make	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	135	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	135	10	as-	as-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	11	sumption	sumption	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	135	12	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	135	13	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	135	15	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	135	16	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	17	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	135	18	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	135	19	clustered	cluster	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	135	20	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	135	21	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	22	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	135	23	terms	term	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	135	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	136	1	Topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	136	2	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	136	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	136	4	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	136	5	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	136	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	136	7	discarded	discard	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	136	8	altogether	altogether	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	136	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	137	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	137	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	137	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	137	4	manual	manual	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	137	5	an-	an-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	137	6	notations	notation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	137	7	available	available	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	137	8	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	137	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	137	10	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	137	11	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	137	12	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	137	13	de-	de-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	137	14	termine	termine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	137	15	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	137	16	clusters	cluster	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	137	17	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	137	18	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	137	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	138	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	2	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	138	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	138	5	100	100	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	138	6	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	138	7	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	138	9	topic	topic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	138	10	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	138	12	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	138	13	examine	examine	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	138	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	138	15	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	138	16	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	138	17	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	138	18	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	19	highest	high	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	138	20	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	21	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	22	that	that	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	138	23	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	138	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	139	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	139	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	139	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	4	precision	precision	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	6	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	7	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	8	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	10	pro-	pro-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	11	portion	portion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	13	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	14	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	139	15	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	139	16	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	17	match	match	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	139	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	19	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	139	20	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	22	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	139	23	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	139	24	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	139	25	1,966	1,966	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	139	26	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	139	27	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	139	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	139	29	cutoff	cutoff	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	30	men-	men-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	139	31	tioned	tione	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	139	32	above	above	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	139	33	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	139	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	140	1	Only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	140	2	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	140	3	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	140	4	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	140	5	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	140	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	140	7	pre-	pre-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	8	cision	cision	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	10	at	at	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	140	11	least	least	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	140	12	25	25	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	140	13	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	14	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	140	15	retained	retain	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	140	17	resulting	result	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	140	18	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	140	19	68	68	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	140	20	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	140	21	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	140	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	141	1	Some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	141	2	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	141	3	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	141	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	141	5	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	141	6	manually	manually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	141	7	created	create	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	141	8	headings	heading	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	141	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	141	10	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	141	11	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	141	12	landscape	landscape	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	141	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	142	1	grass	grass	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	3	mountain	mountain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	5	green	green	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	7	hill	hill	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	9	blue	blue	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	11	field	field	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	13	brown	brown	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	15	sand	sand	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	17	desert	desert	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	19	dirt	dirt	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	21	landscape	landscape	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	142	23	sky	sky	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	142	24	Gilhooly	Gilhooly	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	25	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	142	26	Logie	Logie	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	27	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	142	28	Bristol	Bristol	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	29	Norms	Norms	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	142	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	143	1	3http://nlp.stanford.edu/software/tmt/tmt-0.4/	3http://nlp.stanford.edu/software/tmt/tmt-0.4/	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	143	2	344	344	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	143	3	jewellery	jewellery	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	143	4	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	144	1	silver	silver	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	3	white	white	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	144	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	5	diamond	diamond	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	7	gold	gold	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	9	necklace	necklace	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	11	chain	chain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	13	ring	ring	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	15	jewel	jewel	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	17	wedding	wedding	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	19	diamonds	diamond	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	144	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	144	21	jewelry	jewelry	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	144	22	shapes	shape	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	144	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	145	1	round	round	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	145	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	3	ball	ball	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	5	circles	circle	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	7	logo	logo	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	9	dots	dot	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	11	square	square	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	13	dot	dot	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	15	sphere	sphere	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	17	glass	glass	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	19	hole	hole	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	21	oval	oval	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	23	circle	circle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	24	Combining	combine	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	145	25	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	145	26	68	68	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	145	27	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	145	29	there	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	145	30	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	145	31	5,347	5,347	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	145	32	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	145	33	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	145	34	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	35	because	because	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	36	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	37	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	145	38	overlap	overlap	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	145	39	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	40	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	145	41	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	42	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	145	43	most	most	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	44	probable	probable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	145	45	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	145	46	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	146	1	2,832	2,832	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	146	2	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	146	3	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	146	4	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	146	5	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	146	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	146	7	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	146	8	present	present	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	146	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	146	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	146	11	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	146	12	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	146	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	147	1	Some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	147	2	examples	example	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	147	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	4	new	new	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	147	5	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	147	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	7	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	147	8	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	9	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	147	10	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	11	daffodil	daffodil	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	147	12	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	14	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	15	sailor	sailor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	16	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	18	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	19	hel-	hel-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	147	20	met	meet	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	147	21	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	23	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	24	postcard	postcard	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	147	25	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	26	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	27	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	28	sticker	sticker	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	29	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	31	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	32	carousel	carousel	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	33	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	34	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	35	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	36	kayak	kayak	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	147	37	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	38	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	147	39	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	147	40	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	147	41	camouflage	camouflage	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	147	42	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	147	43	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	148	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	2	later	later	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	148	3	experiments	experiment	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	148	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	148	5	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	148	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	148	7	5,347	5,347	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	148	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	148	9	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	148	11	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	148	12	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	13	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	148	15	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	148	16	keep	keep	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	148	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	148	18	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	19	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	148	21	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	148	22	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	148	23	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	148	24	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	25	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	148	26	68	68	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	148	27	selected	select	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	148	28	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	148	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	149	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	149	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	149	3	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	149	4	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	149	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	149	6	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	149	7	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	149	8	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	149	9	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	149	10	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	149	11	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	149	12	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	149	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	149	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	149	15	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	149	16	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	149	17	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	149	18	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	149	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	150	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	150	2	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	150	3	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	150	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	150	5	adjectives	adjective	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	150	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	150	7	adverbs	adverb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	150	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	150	9	verbs	verb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	150	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	150	11	common	common	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	150	12	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	150	13	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	14	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	150	15	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	16	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	17	candidate	candidate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	18	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	150	19	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	20	computing	compute	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	150	21	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	150	22	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	150	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	151	1	Overall	overall	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	151	2	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	151	3	use	use	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	151	5	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	151	6	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	151	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	151	8	propor-	propor-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	9	tion	tion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	11	candidate	candidate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	12	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	151	13	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	151	14	match	match	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	151	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	151	16	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	151	17	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	18	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	19	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	151	21	TOTAL	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	151	22	VISUAL	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	151	23	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	151	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	152	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	152	3	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	152	5	proportions	proportion	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	152	6	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	7	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	152	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	152	10	first	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	152	11	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	152	12	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	152	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	152	15	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	17	BEG	BEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	18	VISUAL	VISUAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	19	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	152	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	152	22	last	last	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	152	23	200	200	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	152	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	152	25	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	26	END	END	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	27	VISUAL	VISUAL	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	152	28	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	29	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	152	30	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	152	31	middle	middle	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	152	32	region	region	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	33	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	34	MID	MID	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	35	VISUAL	VISUAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	152	36	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	152	37	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	152	38	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	152	39	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	153	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	153	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	153	3	divide	divide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	153	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	153	5	arti-	arti-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	153	6	cle	cle	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	153	7	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	153	8	five	five	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	153	9	equally	equally	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	153	10	sized	sized	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	153	11	bins	bin	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	153	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	153	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	153	14	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	153	15	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	153	16	bin	bin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	153	17	captures	capture	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	153	18	consecutive	consecutive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	153	19	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	153	20	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	153	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	153	22	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	153	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	154	1	Within	within	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	154	2	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	154	3	bin	bin	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	154	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	154	5	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	154	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	154	7	proportion	proportion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	154	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	154	9	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	154	10	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	154	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	155	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	155	2	treat	treat	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	155	3	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	155	4	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	155	5	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	155	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	155	7	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	155	8	distribution	distribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	155	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	155	10	compute	compute	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	155	11	its	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	155	12	entropy	entropy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	155	13	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	155	14	ENTROPY	entropy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	155	15	VISUAL	VISUAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	155	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	155	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	156	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	156	2	expected	expect	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	156	3	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	156	4	position	position	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	5	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	156	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	156	7	indicate	indicate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	156	8	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	156	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	156	10	placement	placement	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	12	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	156	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	156	14	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	156	15	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	16	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	17	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	156	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	157	1	Topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	2	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	157	3	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	157	5	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	157	6	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	157	7	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	157	8	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	157	9	what	what	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	157	10	pro-	pro-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	11	portion	portion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	157	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	157	15	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	157	16	identify	identify	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	157	17	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	18	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	157	19	matches	match	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	157	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	157	21	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	22	under	under	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	23	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	157	24	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	157	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	158	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	158	2	maximum	maximum	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	158	3	proportion	proportion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	158	4	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	158	5	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	158	6	single	single	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	158	7	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	158	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	158	9	MAX	MAX	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	158	10	TOPIC	TOPIC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	158	11	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	158	12	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	158	13	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	158	14	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	158	15	fea-	fea-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	158	16	ture	ture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	158	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	159	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	159	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	159	3	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	159	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	159	5	greedy	greedy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	159	6	cover	cover	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	7	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	9	top-	top-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	159	10	ics	ics	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	11	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	159	13	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	159	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	159	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	159	17	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	159	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	160	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	160	2	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	3	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	160	4	matches	match	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	160	6	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	160	7	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	160	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	10	added	add	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	11	first	first	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	160	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	160	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	160	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	160	15	next	next	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	16	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	17	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	18	selected	select	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	19	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	20	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	160	22	remain-	remain-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	160	23	ing	ing	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	24	unmatched	unmatched	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	160	25	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	160	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	161	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	161	2	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	4	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	161	5	needed	need	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	161	7	cover	cover	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	161	8	50	50	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	161	9	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	161	12	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	13	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	161	14	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	161	15	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	161	16	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	161	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	161	18	TOPIC	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	19	COVER	cover	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	20	VISUAL	visual	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	21	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	161	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	162	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	162	2	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	162	3	cap-	cap-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	162	4	ture	ture	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	162	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	162	6	mix	mix	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	162	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	162	8	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	162	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	162	10	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	162	11	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	162	12	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	162	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	163	1	Disregarding	disregard	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	163	2	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	3	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	163	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	163	6	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	163	7	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	163	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	9	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	10	NUM	NUM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	163	11	PICTURES	PICTURES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	163	12	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	13	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	163	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	15	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	17	im-	im-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	163	18	ages	age	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	163	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	21	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	22	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	163	23	40	40	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	163	24	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	27	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	28	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	163	29	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	163	30	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	163	31	matched	match	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	32	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	33	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	163	34	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	163	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	164	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	164	2	found	find	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	164	3	8	8	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	164	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	164	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	164	6	vary	vary	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	164	7	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	164	8	be-	be-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	164	9	tween	tween	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	164	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	164	11	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	164	12	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	164	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	164	14	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	164	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	165	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	165	2	fea-	fea-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	165	3	tures	ture	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	4	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	165	5	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	165	6	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	165	7	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	165	9	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	165	10	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	165	11	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	12	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	165	13	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	165	14	BEG	BEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	165	15	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	165	17	END	END	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	165	18	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	165	20	MAX	MAX	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	165	21	TOPIC	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	165	22	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	165	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	166	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	166	2	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	3	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	4	signifi-	signifi-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	5	cantly	cantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	166	6	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	166	7	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	166	10	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	166	11	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	12	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	166	13	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	166	14	TOTAL	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	166	15	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	166	17	MID	mid	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	18	VISUAL	VISUAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	166	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	166	20	ENTROPY	entropy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	166	21	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	166	23	TOPIC	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	24	COVER	cover	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	166	25	VISUAL	VISUAL	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	166	26	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	166	27	NUM	NUM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	166	28	PICTURES	PICTURES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	166	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	167	1	It	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	167	2	appears	appear	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	3	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	167	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	167	5	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	6	expectation	expectation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	167	7	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	167	8	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	167	9	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	10	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	167	11	contain	contain	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	167	12	more	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	167	13	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	167	15	overall	overall	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	167	16	does	do	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	17	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	167	18	hold	hold	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	167	19	true	true	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	167	20	here	here	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	167	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	168	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	168	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	168	3	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	168	4	writ-	writ-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	5	ing	ing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	6	samples	sample	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	168	7	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	168	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	168	9	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	168	10	degree	degree	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	12	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	168	13	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	14	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	168	16	beginning	beginning	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	168	18	ends	end	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	168	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	168	20	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	168	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	169	1	Good	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	169	2	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	169	3	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	169	4	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	169	5	lower	low	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	169	6	entropy	entropy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	169	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	169	9	distribution	distribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	11	vi-	vi-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	169	12	sual	sual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	169	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	169	14	indicating	indicate	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	169	15	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	16	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	169	17	appear	appear	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	169	18	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	19	localized	localized	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	169	20	positions	position	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	169	21	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	22	contrast	contrast	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	23	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	24	being	be	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	169	25	distributed	distribute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	26	throughout	throughout	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	169	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	170	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	170	2	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	170	4	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	170	6	further	further	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	170	7	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	170	8	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	170	11	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	170	12	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	170	13	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	170	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	170	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	170	16	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	170	17	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	170	18	come	come	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	170	19	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	20	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	170	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	170	22	few	few	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	170	23	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	170	24	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	170	25	compared	compare	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	26	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	27	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	170	28	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	170	29	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	170	30	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	170	31	so	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	170	32	may	may	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	170	33	evoke	evoke	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	170	34	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	170	35	coherent	coherent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	170	36	image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	37	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	170	38	scene	scene	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	170	39	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	171	1	3.2	3.2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	171	2	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	3	use	use	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	5	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	8	story	story	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	9	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	171	10	hypothesized	hypothesize	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	171	11	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	13	containing	contain	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	171	14	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	15	findings	finding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	16	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	17	directly	directly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	171	18	affect	affect	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	171	19	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	20	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	21	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	22	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	23	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	171	24	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	171	25	therefore	therefore	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	171	26	involve	involve	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	171	27	explicit	explicit	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	171	28	references	reference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	29	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	30	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	171	31	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	32	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	33	story	story	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	34	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	171	35	will	will	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	171	36	make	make	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	171	37	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	38	bigger	big	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	171	39	impact	impact	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	40	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	41	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	171	42	reader	reader	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	171	43	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	172	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	2	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	172	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	172	5	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	172	6	frequent	frequent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	172	7	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	8	among	among	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	9	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	172	10	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	172	11	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	172	12	samples	sample	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	172	13	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	172	14	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	172	15	medicine	medicine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	16	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	172	17	health	health	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	18	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	172	19	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	172	20	ar-	ar-	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	172	21	ticles	ticle	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	172	22	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	172	23	often	often	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	172	24	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	25	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	172	27	view	view	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	28	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	29	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	172	30	patient	patient	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	172	32	doctor	doctor	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	33	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	172	34	scientist	scientist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	172	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	173	1	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	173	2	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	173	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	173	4	below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	173	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	174	1	Dr.	Dr.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	2	Remington	Remington	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	3	was	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	174	4	born	bear	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	6	Reedville	Reedville	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	174	8	Va.	Virginia	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	174	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	11	1922	1922	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	174	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	174	13	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	14	Maud	Maud	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	174	16	P.	P.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	17	Sheldon	Sheldon	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	18	Remington	Remington	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	174	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	174	20	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	174	21	school	school	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	22	headmaster	headmaster	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	174	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	175	1	Charles	Charles	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	175	2	spent	spend	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	175	3	his	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	175	4	boyhood	boyhood	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	175	5	chasing	chase	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	175	6	butterflies	butterfly	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	175	7	alongside	alongside	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	175	8	his	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	175	9	father	father	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	175	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	175	11	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	175	12	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	175	13	col-	col-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	175	14	lector	lector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	175	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	176	1	During	during	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	2	his	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	176	3	graduate	graduate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	176	4	studies	study	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	176	5	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	6	Harvard	Harvard	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	176	8	he	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	9	founded	found	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	176	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	176	11	Lepidopterists	Lepidopterists	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	176	12	’	’	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	176	13	Society	Society	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	14	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	15	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	176	16	equally	equally	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	176	17	butterfly	butterfly	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	18	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	176	19	smitten	smite	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	20	under-	under-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	21	graduate	graduate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	176	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	176	23	Harry	Harry	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	24	Clench	Clench	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	176	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	177	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	177	2	approximate	approximate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	177	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	177	4	facet	facet	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	5	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	6	computing	compute	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	177	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	177	8	num-	num-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	177	9	ber	ber	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	11	explicit	explicit	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	177	12	references	reference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	177	13	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	14	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	177	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	177	16	relying	rely	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	177	17	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	18	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	177	19	sources	source	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	177	20	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	21	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	22	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	23	animacy	animacy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	24	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	177	25	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	177	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	178	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	178	2	first	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	178	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	178	4	named	name	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	5	entity	entity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	6	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	7	NE	NE	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	8	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	9	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	178	10	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	11	PERSON	person	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	178	13	ORGANI-	ORGANI-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	14	ZATION	ZATION	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	178	16	LOCATION	LOCATION	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	17	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	18	returned	return	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	19	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	178	21	Stanford	Stanford	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	22	NE	NE	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	23	recognition	recognition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	24	tool	tool	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	178	25	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	26	Finkel	Finkel	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	27	et	et	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	178	28	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	29	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	178	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	178	31	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	178	32	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	178	33	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	179	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	179	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	179	3	created	create	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	179	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	179	5	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	179	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	179	7	personal	personal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	179	8	pronouns	pronoun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	179	9	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	179	10	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	179	11	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	179	12	he	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	179	13	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	179	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	179	15	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	179	16	my-	my-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	179	17	self	self	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	179	18	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	179	19	etc	etc	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	179	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	180	1	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	180	2	standardly	standardly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	180	3	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	180	4	animate	animate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	180	5	entities	entity	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	180	6	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	180	7	animate	animate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	180	8	pronouns	pronoun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	180	9	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	180	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	181	1	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	181	2	third	third	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	3	resource	resource	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	4	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	181	6	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	8	times	time	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	181	9	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	10	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	11	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	181	12	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	181	13	NP	NP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	181	14	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	181	15	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	181	16	followed	follow	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	17	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	18	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	181	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	181	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	181	21	relative	relative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	22	pronouns	pronoun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	181	23	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	181	24	who	who	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	181	25	’	’	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	181	26	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	181	27	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	181	28	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	181	29	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	181	30	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	181	31	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	181	32	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	181	33	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	181	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	182	1	345	345	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	182	2	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	182	3	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	182	4	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	182	5	664,673	664,673	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	182	6	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	182	7	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	182	8	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	182	9	col-	col-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	182	10	lected	lecte	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	182	11	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	182	12	Ji	Ji	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	182	14	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	15	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	182	16	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	182	17	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	182	18	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	182	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	182	20	Google	Google	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	21	Ngram	Ngram	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	22	Corpus	Corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	23	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	182	24	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	25	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	26	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	27	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	182	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	182	29	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	182	30	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	182	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	183	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	183	4	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	183	5	heuris-	heuris-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	6	tic	tic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	7	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	183	8	obtain	obtain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	183	9	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	183	10	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	12	animate	animate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	13	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	183	14	google	google	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	15	animate	animate	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	183	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	183	18	inanimate	inanimate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	183	19	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	183	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	183	21	google	google	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	22	inanimate	inanimate	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	183	23	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	183	24	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	25	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	183	26	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	183	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	184	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	184	2	head	head	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	4	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	184	5	NP	NP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	184	6	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	184	7	taken	take	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	8	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	9	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	184	10	candidate	candidate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	11	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	184	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	185	1	If	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	185	3	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	4	does	do	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	185	5	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	185	6	occur	occur	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	185	7	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	8	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	185	9	who	who	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	185	10	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	185	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	12	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	185	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	185	15	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	185	16	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	185	17	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	185	18	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	185	19	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	185	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	185	21	head	head	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	185	23	then	then	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	185	24	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	185	25	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	185	26	inani-	inani-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	27	mate	mate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	185	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	186	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	186	2	contrast	contrast	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	186	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	186	4	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	186	5	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	186	6	appears	appear	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	186	7	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	186	8	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	186	9	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	186	10	who	who	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	186	11	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	186	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	186	13	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	186	14	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	186	15	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	186	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	186	17	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	186	18	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	186	19	animate	animate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	186	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	187	1	Otherwise	otherwise	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	187	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	187	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	4	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	5	NP	NP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	6	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	187	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	8	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	187	10	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	11	head	head	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	187	13	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	14	check	check	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	15	whether	whether	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	17	count	count	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	18	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	19	times	time	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	187	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	21	noun	noun	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	22	phrase	phrase	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	23	appeared	appear	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	187	24	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	25	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	187	26	who	who	WP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	27	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	187	28	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	187	29	greater	great	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	187	30	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	31	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	32	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	33	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	34	occurrences	occurrence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	187	35	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	36	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	187	37	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	38	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	187	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	187	40	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	187	41	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	187	42	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	187	43	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	187	44	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	187	45	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	187	46	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	187	47	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	187	48	taken	take	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	49	individually	individually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	187	50	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	187	51	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	187	52	that	that	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	187	53	noun	noun	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	54	phrase	phrase	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	187	55	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	188	1	If	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	188	3	condition	condition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	4	holds	hold	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	188	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	6	at	at	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	188	7	least	least	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	188	8	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	188	9	noun	noun	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	188	10	phrase	phrase	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	188	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	188	13	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	14	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	188	15	marked	mark	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	16	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	188	17	animate	animate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	188	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	189	1	When	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	189	2	computing	compute	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	189	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	189	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	189	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	189	6	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	189	7	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	189	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	189	9	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	189	10	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	189	11	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	189	12	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	189	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	189	14	pronouns	pronoun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	189	15	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	189	16	candidate	candidate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	189	17	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	189	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	190	1	If	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	190	3	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	4	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	190	5	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	190	6	pronoun	pronoun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	190	8	appears	appear	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	190	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	10	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	190	11	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	13	an-	an-	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	190	14	imate	imate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	190	15	pronouns	pronoun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	190	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	190	17	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	190	18	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	190	19	assigned	assign	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	20	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	190	21	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	190	22	animate	animate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	190	23	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	190	24	label	label	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	190	25	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	190	26	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	190	27	inanimate	inanimate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	190	28	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	190	29	otherwise	otherwise	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	190	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	191	1	If	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	191	3	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	4	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	191	5	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	191	6	proper	proper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	191	7	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	191	9	tagged	tag	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	191	10	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	191	12	PERSON	person	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	13	NE	NE	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	14	tag	tag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	191	16	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	17	mark	mark	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	18	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	19	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	20	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	191	21	animate	animate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	191	22	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	191	23	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	191	24	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	25	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	26	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	191	27	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	191	28	ORGANIZATION	ORGANIZATION	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	29	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	191	30	LO-	LO-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	31	CATION	CATION	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	191	32	tag	tag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	33	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	191	34	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	191	35	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	191	36	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	191	37	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	191	38	inanimate	inanimate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	191	39	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	191	40	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	192	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	192	2	common	common	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	192	3	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	192	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	192	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	192	6	check	check	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	192	7	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	192	8	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	192	9	appears	appear	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	192	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	192	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	192	12	google	google	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	192	13	animate	animate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	192	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	192	15	inanimate	inanimate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	192	16	lists	list	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	192	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	193	1	Any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	193	2	match	match	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	193	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	193	4	labelled	label	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	193	5	accord-	accord-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	193	6	ingly	ingly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	193	7	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	193	8	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	193	9	animate	animate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	193	10	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	193	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	193	12	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	193	13	inanimate	inanimate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	193	14	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	193	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	194	1	Note	note	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	194	2	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	194	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	194	4	procedure	procedure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	194	5	may	may	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	194	6	leave	leave	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	194	7	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	194	8	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	194	9	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	194	10	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	194	11	labels	label	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	194	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	195	1	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	195	2	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	4	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	6	animate	animate	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	7	tokens	token	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	9	ANIM	ANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	10	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	195	12	inanimate	inanimate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	195	13	tokens	token	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	15	INAMIN	INAMIN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	195	18	both	both	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	195	19	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	195	20	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	21	normalized	normalize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	22	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	23	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	195	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	195	25	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	195	27	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	195	28	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	29	ANIM	ANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	30	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	195	32	INANIM	INANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	33	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	195	34	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	195	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	196	1	Three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	196	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	196	3	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	196	4	fea-	fea-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	196	5	tures	ture	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	196	6	had	have	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	196	7	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	196	8	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	196	9	mean	mean	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	196	10	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	196	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	196	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	196	13	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	14	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	196	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	196	16	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	196	17	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	196	18	ANIM	ANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	196	20	ANIM	ANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	21	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	196	23	INANIM	INANIM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	24	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	196	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	197	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	2	found	find	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	197	3	upon	upon	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	4	observation	observation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	5	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	197	6	sev-	sev-	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	197	7	eral	eral	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	197	8	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	197	9	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	197	10	talk	talk	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	11	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	12	government	government	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	13	policies	policy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	197	14	in-	in-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	15	volve	volve	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	16	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	197	17	lot	lot	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	18	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	19	references	reference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	197	20	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	21	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	197	22	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	197	23	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	24	often	often	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	197	25	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	197	27	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	197	28	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	197	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	198	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	198	2	findings	finding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	198	3	suggest	suggest	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	198	4	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	198	6	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	198	7	human	human	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	198	8	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	198	9	dimension	dimension	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	10	might	may	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	198	11	need	need	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	198	12	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	198	13	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	198	14	computed	compute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	15	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	198	16	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	198	17	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	18	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	19	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	198	20	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	198	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	22	references	reference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	198	23	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	24	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	198	25	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	198	26	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	198	27	involve	involve	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	198	28	finer	fine	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	198	29	distinctions	distinction	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	198	30	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	198	31	them	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	198	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	199	1	3.3	3.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	199	2	Beautiful	beautiful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	199	3	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	4	Beautiful	beautiful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	199	5	phrasing	phrasing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	199	7	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	8	choice	choice	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	9	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	199	10	entertain	entertain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	199	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	199	12	reader	reader	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	199	14	leave	leave	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	199	15	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	199	16	positive	positive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	199	17	impression	impression	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	199	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	200	1	Multi-	Multi-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	2	ple	ple	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	3	studies	study	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	4	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	200	6	education	education	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	7	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	200	9	Diederich	Diederich	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	200	11	1974	1974	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	200	12	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	200	13	Spandel	Spandel	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	200	15	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	200	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	200	17	note	note	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	18	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	19	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	200	20	teachers	teacher	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	21	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	200	22	expert	expert	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	200	23	adult	adult	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	24	readers	reader	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	25	graded	grade	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	200	26	student	student	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	27	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	200	29	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	30	choice	choice	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	31	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	200	32	phrasing	phrasing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	33	always	always	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	200	34	turn	turn	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	35	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	200	36	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	200	37	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	200	38	significant	significant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	200	39	factors	factor	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	40	influencing	influence	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	200	41	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	200	42	raters	rater	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	43	’	’	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	44	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	200	45	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	201	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	201	2	implement	implement	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	201	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	4	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	6	detecting	detect	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	201	7	creative	creative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	8	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	9	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	10	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	11	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	12	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	13	idea	idea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	14	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	15	creative	creative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	16	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	201	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	201	18	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	201	19	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	201	20	sometimes	sometimes	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	201	21	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	22	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	23	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	201	24	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	25	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	26	un-	un-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	27	usual	usual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	28	contexts	contexts	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	201	29	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	201	30	combinations	combination	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	201	31	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	201	32	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	33	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	201	34	sound	sound	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	201	35	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	201	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	202	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	202	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	202	3	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	202	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	5	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	202	6	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	7	both	both	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	202	8	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	202	10	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	12	individual	individual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	202	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	202	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	202	15	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	202	17	com-	com-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	18	bination	bination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	20	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	202	21	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	22	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	202	23	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	202	24	relation	relation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	202	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	203	1	Word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	2	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	3	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	203	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	203	5	Unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	203	6	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	203	7	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	8	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	203	9	ar-	ar-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	10	ticle	ticle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	11	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	203	12	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	203	13	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	203	14	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	203	15	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	203	16	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	17	low	low	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	203	18	frequencies	frequency	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	203	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	20	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	203	21	background	background	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	22	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	203	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	204	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	204	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	204	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	204	4	full	full	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	204	5	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	7	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	204	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	204	9	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	204	10	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	204	11	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	12	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	204	13	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	14	year	year	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	15	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	204	16	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	204	18	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	204	19	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	20	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	204	22	background	background	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	23	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	204	24	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	204	25	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	204	26	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	204	27	over-	over-	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	204	28	lap	lap	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	204	29	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	30	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	204	31	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	32	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	33	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	34	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	204	35	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	204	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	205	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	205	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	205	3	ex-	ex-	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	205	4	plore	plore	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	205	5	patterns	pattern	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	7	letters	letter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	205	9	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	10	sequences	sequence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	11	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	205	13	idea	idea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	14	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	15	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	205	16	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	205	18	characters	character	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	205	20	phonemes	phoneme	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	21	could	could	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	205	22	create	create	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	205	23	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	205	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	205	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	206	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	206	2	used	use	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	206	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	206	4	CMU	CMU	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	206	5	pronunciation	pronunciation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	6	dictionary	dictionary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	206	7	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	206	8	Weide	Weide	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	206	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	206	10	1998	1998	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	206	11	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	206	12	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	206	13	get	get	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	206	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	206	15	phoneme	phoneme	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	206	16	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	18	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	206	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	206	20	built	build	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	206	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	206	22	4-	4-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	206	23	gram	gram	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	24	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	26	phonemes	phoneme	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	206	27	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	206	29	background	background	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	30	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	206	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	207	1	Laplace	Laplace	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	207	2	smoothing	smooth	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	207	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	207	4	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	207	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	207	6	compute	compute	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	207	7	probabilities	probability	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	207	8	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	207	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	207	10	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	207	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	208	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	208	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	208	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	208	4	CMU	CMU	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	208	5	dictionary	dictionary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	6	does	do	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	208	7	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	208	8	contain	contain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	208	9	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	10	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	11	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	12	several	several	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	208	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	208	14	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	15	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	208	16	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	208	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	209	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	209	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	3	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	209	4	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	5	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	209	6	approximate	approximate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	209	7	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	8	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	209	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	209	10	letters	letter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	209	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	209	15	obtain	obtain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	209	16	an-	an-	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	209	17	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	209	18	4-gram	4-gram	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	209	19	model.4	model.4	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	20	Only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	209	21	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	22	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	209	23	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	24	longer	long	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	209	25	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	26	4	4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	209	27	characters	character	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	28	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	29	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	30	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	31	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	209	32	models	model	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	33	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	209	34	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	35	fil-	fil-	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	209	36	ter	ter	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	209	37	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	209	38	proper	proper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	209	39	names	name	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	40	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	209	41	named	name	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	209	42	entities	entity	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	43	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	209	44	numbers	number	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	209	45	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	210	1	During	during	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	2	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	210	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	210	5	analyzed	analyze	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	210	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	210	7	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	210	8	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	9	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	210	10	entire	entire	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	210	11	year	year	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	13	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	210	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	210	15	1997	1997	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	210	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	210	17	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	210	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	210	19	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	210	20	models	model	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	210	21	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	210	22	identify	identify	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	210	23	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	210	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	210	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	211	1	Below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	211	2	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	211	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	211	4	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	6	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	211	7	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	8	lowest	low	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	211	9	frequency	frequency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	211	11	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	211	12	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	13	high-	high-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	14	est	est	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	211	15	perplexity	perplexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	16	under	under	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	211	18	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	211	20	letter	letter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	211	21	models	model	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	211	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	212	1	Low	low	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	212	2	frequency	frequency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	212	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	213	1	undersheriff	undersheriff	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	213	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	3	woggle	woggle	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	213	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	5	ahmok	ahmok	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	213	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	7	hofman	hofman	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	213	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	9	volga	volga	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	213	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	11	oceanaut	oceanaut	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	13	trachoma	trachoma	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	15	baneful	baneful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	213	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	17	truffler	truffler	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	19	acrimal	acrimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	213	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	21	corvair	corvair	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	213	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	213	23	entomopter	entomopter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	24	High	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	213	25	perplexity	perplexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	26	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	213	27	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	28	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	213	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	214	1	showroom	showroom	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	3	yahoo	yahoo	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	5	dossier	dossier	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	7	powwow	powwow	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	9	plowshare	plowshare	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	11	oomph	oomph	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	13	chihuahua	chihuahua	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	15	iono-	iono-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	16	sphere	sphere	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	18	boudoir	boudoir	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	20	superb	superb	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	22	zaire	zaire	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	214	23	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	214	24	oeuvre	oeuvre	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	214	25	High	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	214	26	perplexity	perplexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	27	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	214	28	letter	letter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	29	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	214	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	215	1	kudzu	kudzu	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	3	muumuu	muumuu	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	5	qi-	qi-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	6	pao	pao	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	8	yugoslav	yugoslav	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	10	kohlrabi	kohlrabi	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	12	iraqi	iraqi	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	215	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	14	yaqui	yaqui	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	16	yakuza	yakuza	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	18	jujitsu	jujitsu	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	20	oeu-	oeu-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	21	vre	vre	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	23	yaohan	yaohan	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	24	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	25	kaffiyeh	kaffiyeh	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	26	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	215	27	computing	compute	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	215	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	215	29	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	215	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	31	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	32	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	215	33	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	215	34	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	215	35	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	36	verbs	verb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	215	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	215	38	adjectives	adjective	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	215	39	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	215	40	adverbs	adverb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	215	41	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	216	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	216	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	216	3	require	require	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	216	4	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	216	6	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	216	7	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	216	8	at	at	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	9	least	least	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	216	10	5	5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	216	11	letters	letter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	216	12	long	long	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	216	13	4We	4we	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	216	14	found	find	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	216	15	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	16	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	216	17	order	order	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	18	n	n	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	19	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	216	20	grams	gram	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	216	21	provided	provide	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	22	better	well	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	216	23	pre-	pre-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	216	24	dictions	diction	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	216	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	26	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	216	27	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	28	during	during	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	29	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	216	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	217	1	346	346	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	217	2	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	217	3	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	217	4	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	217	5	contain	contain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	217	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	217	7	hyphen5	hyphen5	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	217	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	218	1	Three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	218	2	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	218	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	218	4	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	218	5	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	218	6	computed	compute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	218	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	219	1	FREQ	freq	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	219	2	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	219	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	219	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	219	5	aver-	aver-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	219	6	age	age	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	219	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	219	8	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	219	9	frequencies	frequency	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	219	10	computed	compute	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	219	11	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	219	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	219	13	back-	back-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	219	14	ground	ground	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	219	15	corpus	corpus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	219	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	220	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	220	2	second	second	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	220	3	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	220	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	220	7	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	220	10	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	11	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	220	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	221	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	221	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	221	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	221	4	average	average	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	221	5	perplexity	perplexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	221	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	221	7	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	221	8	under	under	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	221	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	221	10	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	221	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	221	12	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	221	13	PHONEME	PHONEME	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	221	14	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	221	15	ALL	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	221	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	222	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	2	addition	addition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	222	4	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	222	5	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	222	6	or-	or-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	222	7	der	der	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	222	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	222	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	11	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	12	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	13	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	14	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	15	decreasing	decrease	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	222	16	per-	per-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	222	17	plexity	plexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	18	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	222	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	222	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	21	average	average	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	222	22	perplexity	perplexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	23	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	24	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	25	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	222	26	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	27	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	222	28	20	20	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	29	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	222	30	30	30	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	31	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	222	32	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	33	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	34	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	35	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	222	36	added	add	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	37	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	222	38	fea-	fea-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	222	39	tures	ture	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	222	40	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	222	41	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	222	42	PHONEME	PHONEME	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	222	43	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	222	44	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	45	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	222	46	20	20	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	47	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	222	48	30	30	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	222	49	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	222	50	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	223	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	2	ob-	ob-	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	3	tain	tain	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	4	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	223	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	223	6	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	223	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	223	8	letter	letter	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	223	9	n	n	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	223	10	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	223	11	gram	gram	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	223	12	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	223	13	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	223	14	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	15	CHAR	CHAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	16	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	17	ALL	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	223	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	223	19	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	20	CHAR	CHAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	21	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	223	22	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	223	23	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	223	24	20	20	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	223	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	223	26	30	30	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	223	27	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	223	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	224	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	224	2	phoneme	phoneme	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	224	3	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	224	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	224	5	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	224	6	ignore	ignore	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	224	7	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	224	8	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	224	9	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	224	10	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	224	11	have	have	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	224	12	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	224	13	entry	entry	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	224	14	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	224	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	224	16	CMU	CMU	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	224	17	dictionary	dictionary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	224	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	225	1	Word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	225	2	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	225	3	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	225	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	225	5	Next	next	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	225	6	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	225	7	attempt	attempt	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	225	8	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	225	9	detect	detect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	225	10	un-	un-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	225	11	usual	usual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	225	12	combinations	combination	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	225	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	225	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	225	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	226	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	226	2	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	226	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	226	4	calculation	calculation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	5	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	226	6	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	7	certain	certain	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	8	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	10	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	11	relations	relation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	12	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	226	13	a	a	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	14	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	226	15	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	16	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	226	17	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	226	18	adjective	adjective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	19	modifiers	modifier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	226	21	b	b	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	226	23	verbs	verb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	24	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	25	adverb	adverb	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	26	modifiers	modifiers	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	226	27	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	226	28	c	c	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	29	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	226	30	adjacent	adjacent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	31	nouns	noun	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	32	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	33	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	226	34	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	35	phrase	phrase	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	36	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	226	37	d	d	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	38	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	226	39	verb	verb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	226	40	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	226	41	subject	subject	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	226	42	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	226	43	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	227	1	Counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	227	2	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	227	3	co	co	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	227	4	-	-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	227	5	occurrence	occurrence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	227	6	again	again	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	227	7	come	come	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	227	8	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	227	9	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	227	10	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	227	11	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	227	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	228	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	228	2	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	228	3	relations	relation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	228	4	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	5	obtained	obtain	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	6	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	228	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	228	8	constituency	constituency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	228	10	de-	de-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	11	pendency	pendency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	12	parses	parse	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	228	13	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	228	15	Stanford	Stanford	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	16	parser	parser	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	228	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	228	18	Klein	Klein	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	228	20	Manning	Manning	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	228	22	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	228	23	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	228	24	De	De	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	25	Marneffe	Marneffe	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	26	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	27	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	28	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	228	29	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	228	30	2006	2006	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	228	31	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	228	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	229	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	229	2	avoid	avoid	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	229	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	229	4	influence	influence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	6	proper	proper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	229	7	names	name	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	229	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	229	9	named	name	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	10	entities	entity	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	229	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	229	12	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	229	13	replace	replace	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	229	14	them	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	229	15	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	16	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	229	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	229	18	NNP	NNP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	229	19	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	20	proper	proper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	229	21	names	name	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	229	22	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	229	23	PERSON	person	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	24	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	229	25	ORG	ORG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	229	26	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	229	27	LOC	loc	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	28	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	29	named	name	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	229	30	entities	entity	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	229	31	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	229	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	230	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	230	2	treat	treat	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	230	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	230	4	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	230	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	6	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	230	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	230	8	dependency	dependency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	9	holds	hold	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	230	10	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	11	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	230	12	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	230	13	auxiliary	auxiliary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	230	14	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	230	16	main	main	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	230	17	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	230	19	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	230	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	231	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	2	adjective	adjective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	231	4	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	231	6	adverb	adverb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	231	8	verb	verb	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	231	9	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	231	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	231	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	231	12	auxiliary	auxiliary	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	13	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	231	15	adjective	adjective	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	16	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	231	17	adverb	adverb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	18	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	231	19	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	20	noun	noun	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	231	21	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	231	22	noun	noun	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	23	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	231	24	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	231	25	it	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	231	26	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	231	28	first	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	29	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	30	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	231	31	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	231	32	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	33	verb	verb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	34	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	231	35	subject	subject	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	36	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	231	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	231	38	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	231	39	auxil-	auxil-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	40	iary	iary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	41	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	231	42	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	231	43	subject	subject	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	231	44	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	232	1	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	232	2	idea	idea	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	232	4	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	232	5	compute	compute	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	232	6	usualness	usualness	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	232	7	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	232	8	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	9	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	10	frequency	frequency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	11	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	12	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	232	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	232	14	particular	particular	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	232	15	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	17	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	232	18	appears	appear	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	232	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	232	21	background	background	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	232	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	233	1	Specifically	specifically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	233	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	233	3	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	233	4	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	233	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	233	6	conditional	conditional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	233	7	proba-	proba-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	8	bility	bility	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	233	11	auxiliary	auxiliary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	233	12	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	13	given	give	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	233	15	main	main	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	233	16	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	17	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	233	19	score	score	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	20	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	21	likelihood	likelihood	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	23	observing	observe	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	233	24	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	233	25	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	233	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	234	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	234	2	consider	consider	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	234	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	234	4	main	main	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	234	5	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	6	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	7	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	8	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	234	10	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	11	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	234	13	so	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	234	14	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	234	15	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	234	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	234	17	conditional	conditional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	234	18	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	20	auxil-	auxil-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	234	21	iary	iary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	234	22	given	give	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	23	main	main	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	234	24	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	25	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	234	26	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	234	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	234	28	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	234	29	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	234	30	around	around	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	234	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	235	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	235	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	235	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	235	4	conditional	conditional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	235	5	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	6	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	235	7	no	no	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	235	8	infor-	infor-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	235	9	mation	mation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	10	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	235	12	frequency	frequency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	13	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	235	15	auxiliary	auxiliary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	235	16	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	235	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	236	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	236	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	236	3	apply	apply	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	236	4	ideas	idea	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	236	5	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	6	interpolation	interpolation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	7	smoothing	smooth	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	236	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	236	9	Chen	Chen	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	236	10	5We	5we	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	236	11	noticed	notice	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	236	12	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	13	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	14	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	236	15	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	16	several	several	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	236	17	new	new	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	236	18	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	236	19	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	236	20	created	create	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	21	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	236	22	hyphen	hyphen	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	236	23	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	236	24	concatenate	concatenate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	236	25	common	common	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	236	26	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	236	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	237	1	ADJ	ADJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	2	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	237	3	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	4	ADV	ADV	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	237	6	VERB	VERB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	7	hypoactive	hypoactive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	8	NNP	NNP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	9	suburbs	suburb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	237	10	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	11	plasticky	plasticky	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	12	woman	woman	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	13	integral	integral	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	14	was	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	15	psychogenic	psychogenic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	16	problems	problem	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	237	17	collective	collective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	18	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	19	yoplait	yoplait	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	20	television	television	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	21	physiologically	physiologically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	237	22	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	23	subminimal	subminimal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	24	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	25	amuck	amuck	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	26	run	run	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	27	ehatchery	ehatchery	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	28	investment	investment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	29	illegitimately	illegitimately	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	237	30	put	put	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	31	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	32	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	237	33	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	34	SUBJ	SUBJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	35	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	237	36	VERB	VERB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	37	specification	specification	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	38	today	today	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	39	blog	blog	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	40	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	41	auditory	auditory	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	42	system	system	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	43	briefer	briefer	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	44	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	45	pal	pal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	46	programs	programs	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	47	hr	hr	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	48	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	49	steganography	steganography	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	50	programs	programs	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	51	knucklehead	knucklehead	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	52	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	53	wastewater	wastewater	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	54	system	system	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	55	lymphedema	lymphedema	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	56	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	57	autism	autism	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	58	conference	conference	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	59	permissions	permission	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	237	60	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	61	Table	table	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	62	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	63	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	237	64	Unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	65	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	66	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	237	67	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	237	68	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	69	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	70	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	237	71	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	237	72	Goodman	Goodman	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	237	73	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	237	74	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	237	75	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	237	76	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	237	77	compute	compute	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	237	78	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	237	79	conditional	conditional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	80	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	81	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	82	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	237	83	interpolated	interpolate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	84	quantity	quantity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	85	together	together	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	237	86	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	87	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	237	88	unigram	unigram	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	89	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	90	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	91	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	237	92	auxiliary	auxiliary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	237	93	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	237	94	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	238	1	p̂(aux|main	p̂(aux|main	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	238	2	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	238	3	=	=	NFP
work_hea3gijzhnc2rl4sxgyj6pvghm	238	4	λ∗p(aux|main)+(1−λ)∗p(aux	λ∗p(aux|main)+(1−λ)∗p(aux	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	238	5	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	238	6	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	238	7	unigram	unigram	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	238	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	238	9	conditional	conditional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	238	10	probabilities	probability	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	238	11	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	238	12	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	238	13	smoothed	smooth	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	238	14	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	238	15	Laplace	Laplace	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	238	16	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	238	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	239	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	2	train	train	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	239	4	lambda	lambda	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	5	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	239	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	239	7	optimize	optimize	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	239	8	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	239	9	likelihood	likelihood	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	10	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	239	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	239	12	Baum	Baum	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	13	Welch	Welch	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	14	algorithm	algorithm	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	239	16	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	239	18	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	239	19	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	20	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	239	21	1997	1997	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	239	22	year	year	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	23	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	239	24	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	25	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	239	26	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	27	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	239	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	240	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	240	2	lambda	lambda	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	3	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	240	4	across	across	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	5	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	240	6	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	240	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	8	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	240	9	tended	tend	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	240	10	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	240	11	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	240	12	lower	low	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	240	13	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	14	0.5	0.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	240	15	giving	give	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	240	16	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	240	17	weight	weight	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	18	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	240	20	unigram	unigram	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	240	21	proba-	proba-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	22	bility	bility	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	23	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	24	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	240	25	auxiliary	auxiliary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	240	26	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	240	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	241	1	Based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	2	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	3	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	241	4	observations	observation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	241	5	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	241	7	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	8	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	241	10	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	241	11	picked	pick	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	241	12	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	241	13	cutoff	cutoff	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	15	0.0001	0.0001	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	241	16	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	241	18	proba-	proba-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	19	bility	bility	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	241	21	0.001	0.001	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	241	22	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	23	adverb	adverb	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	241	24	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	241	25	verb	verb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	26	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	241	27	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	241	28	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	241	29	consider	consider	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	241	30	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	241	31	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	32	probability	probability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	33	below	below	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	34	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	241	35	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	36	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	241	37	un-	un-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	241	38	usual	usual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	241	39	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	242	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	2	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	3	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	4	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	242	6	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	7	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	9	num-	num-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	10	ber	ber	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	12	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	242	13	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	242	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	15	total	total	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	16	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	17	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	18	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	242	19	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	20	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	22	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	23	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	24	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	25	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	242	27	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	242	28	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	29	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	30	normal-	normal-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	242	31	ized	ize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	32	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	33	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	242	34	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	35	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	36	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	37	tokens	token	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	242	38	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	39	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	242	40	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	41	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	42	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	43	WD	WD	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	44	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	45	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	242	46	normalized	normalize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	47	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	48	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	49	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	242	50	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	242	51	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	52	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	53	PH	PH	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	242	54	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	242	55	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	243	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	243	3	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	243	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	6	indi-	indi-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	243	7	vidual	vidual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	243	8	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	9	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	243	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	243	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	12	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	243	13	case	case	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	243	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	243	16	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	18	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	243	19	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	243	20	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	243	21	normalized	normalize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	22	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	243	24	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	243	25	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	243	26	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	243	28	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	243	29	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	243	30	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	31	ADJ	ADJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	32	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	33	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	243	34	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	35	ADV	ADV	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	36	VERB	VERB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	243	38	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	39	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	40	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	41	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	243	42	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	43	SUBJ	SUBJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	44	VERB	VERB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	243	45	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	243	46	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	244	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	244	2	list	list	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	244	5	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	244	6	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	244	7	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	244	8	under	under	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	244	10	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	244	11	pair	pair	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	12	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	244	13	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	244	14	shown	show	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	16	Table	table	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	244	17	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	244	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	245	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	245	2	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	245	3	com-	com-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	4	puted	put	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	5	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	6	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	245	7	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	245	9	random	random	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	245	10	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	245	13	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	14	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	245	15	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	245	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	246	1	Several	several	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	246	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	246	4	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	246	5	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	246	6	involve	involve	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	246	7	hyphenated	hyphenate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	246	9	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	246	10	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	246	11	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	246	12	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	13	themselves	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	246	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	246	15	so	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	246	16	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	246	17	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	246	18	show	show	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	246	19	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	246	21	table	table	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	246	23	top	top	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	246	24	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	246	25	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	246	26	hyphens	hyphen	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	246	27	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	247	1	347	347	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	247	2	Most	Most	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	247	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	247	4	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	247	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	247	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	247	7	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	247	8	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	247	9	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	247	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	247	11	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	247	12	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	247	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	248	1	Those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	248	2	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	248	3	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	248	4	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	248	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	248	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	248	7	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	248	8	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	248	9	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	248	10	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	11	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	248	12	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	13	PHONEME	PHONEME	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	14	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	15	ALL	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	248	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	17	AVR	AVR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	18	CHAR	CHAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	19	PERP	PERP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	248	21	ALL	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	248	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	23	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	248	24	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	248	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	26	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	27	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	28	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	29	PH	PH	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	31	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	32	WD	WD	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	33	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	34	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	35	ADJ	ADJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	36	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	38	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	39	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	40	NOUN	NOUN	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	41	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	248	42	SURP	SURP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	43	SUBJ	SUBJ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	44	VERB	VERB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	248	45	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	249	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	249	2	FREQ	FREQ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	249	3	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	249	4	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	249	5	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	249	6	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	249	7	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	249	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	249	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	249	10	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	249	11	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	249	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	250	1	All	all	PDT
work_hea3gijzhnc2rl4sxgyj6pvghm	250	2	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	250	3	trends	trend	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	250	4	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	250	5	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	250	6	unusual	unusual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	250	7	phrases	phrase	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	250	8	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	250	9	associated	associate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	250	10	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	250	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	250	12	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	250	13	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	250	14	category	category	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	250	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	250	16	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	250	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	251	1	3.4	3.4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	251	2	Sub	Sub	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	251	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	251	4	genres	genre	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	5	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	251	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	251	7	several	several	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	251	8	sub	sub	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	9	-	-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	10	genres	genre	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	11	within	within	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	12	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	13	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	251	15	Stocking	stocking	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	17	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	251	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	251	19	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	251	20	short	short	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	251	21	descriptions	description	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	23	discoveries	discovery	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	24	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	25	longer	long	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	251	26	explanatory	explanatory	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	251	27	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	29	narratives	narrative	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	31	stories	story	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	32	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	33	scientists	scientist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	34	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	35	reports	report	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	36	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	37	meetings	meeting	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	38	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	251	39	review	review	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	40	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	41	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	251	42	blog	blog	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	251	43	posts	post	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	251	44	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	252	1	Naturally	naturally	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	252	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	252	3	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	252	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	252	5	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	252	6	sub	sub	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	252	7	-	-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	252	8	genres	genre	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	252	9	will	will	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	252	10	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	252	11	more	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	252	12	appealing	appealing	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	252	13	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	252	14	readers	reader	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	252	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	253	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	253	2	investigate	investigate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	253	3	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	253	4	aspect	aspect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	253	6	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	253	7	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	253	8	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	253	9	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	10	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	253	11	sub	sub	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	12	-	-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	253	13	genres	genre	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	253	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	15	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	16	—	—	:
work_hea3gijzhnc2rl4sxgyj6pvghm	253	17	narrative	narrative	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	253	19	attribution	attribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	253	21	interview	interview	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	253	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	254	1	Narrative	narrative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	254	2	texts	text	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	254	3	typically	typically	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	254	4	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	254	5	characters	character	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	254	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	254	7	events	event	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	254	8	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	254	9	Nakhimovsky	Nakhimovsky	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	254	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	254	11	1988	1988	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	254	12	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	254	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	254	14	so	so	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	254	15	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	254	16	look	look	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	254	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	254	18	entities	entity	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	254	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	254	20	past	past	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	254	21	tense	tense	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	254	22	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	254	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	254	24	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	254	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	255	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	255	2	count	count	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	255	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	255	4	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	6	sentences	sentence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	255	7	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	255	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	255	9	first	first	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	255	10	verb	verb	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	12	surface	surface	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	13	order	order	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	14	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	255	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	255	17	past	past	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	255	18	tense	tense	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	255	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	256	1	Then	then	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	256	2	among	among	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	3	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	4	sentences	sentence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	256	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	256	6	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	256	7	pick	pick	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	256	8	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	9	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	10	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	256	11	either	either	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	256	12	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	13	personal	personal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	256	14	pronoun	pronoun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	15	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	256	16	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	17	proper	proper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	256	18	noun	noun	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	19	before	before	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	256	21	target	target	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	22	verb	verb	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	256	23	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	256	24	again	again	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	256	25	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	26	surface	surface	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	27	order	order	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	256	28	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	256	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	257	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	257	2	proportion	proportion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	4	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	257	5	sentences	sentence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	257	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	257	8	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	257	10	taken	take	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	11	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	257	13	NARRATIVE	NARRATIVE	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	257	14	score	score	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	257	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	258	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	258	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	258	3	developed	develop	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	258	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	258	5	measure	measure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	258	7	identify	identify	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	258	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	258	9	de-	de-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	10	gree	gree	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	11	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	258	12	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	258	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	258	14	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	15	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	258	16	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	17	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	258	18	attributed	attribute	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	19	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	20	ex-	ex-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	258	21	ternal	ternal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	258	22	sources	source	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	258	23	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	24	opposed	oppose	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	25	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	258	27	author	author	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	258	28	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	258	29	own	own	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	258	30	state-	state-	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	258	31	ments	ment	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	258	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	259	1	Attribution	attribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	2	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	3	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	259	4	sources	source	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	259	5	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	259	6	frequent	frequent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	259	7	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	259	9	news	news	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	10	domain	domain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	11	since	since	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	12	many	many	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	259	13	comments	comment	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	259	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	259	15	opin-	opin-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	259	16	ions	ion	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	259	17	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	259	18	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	259	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	259	20	views	view	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	259	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	259	23	journalist	journalist	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	259	24	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	259	25	Semetko	Semetko	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	259	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	259	27	Valkenburg	Valkenburg	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	259	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	259	29	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	259	30	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	259	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	260	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	2	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	3	news	news	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	260	5	attribution	attribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	6	be-	be-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	260	7	comes	come	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	260	8	more	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	260	9	important	important	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	260	10	since	since	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	260	12	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	13	findings	finding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	260	14	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	260	15	obtained	obtain	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	16	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	17	scientists	scientist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	260	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	260	19	reported	report	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	20	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	21	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	260	22	second-	second-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	23	hand	hand	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	24	manner	manner	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	25	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	260	26	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	260	27	journalists	journalist	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	260	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	261	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	2	ATTRIB	ATTRIB	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	261	3	score	score	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	4	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	261	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	6	proportion	proportion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	8	sentences	sentence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	261	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	11	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	12	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	13	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	261	14	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	15	quote	quote	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	16	symbol	symbol	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	261	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	261	18	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	261	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	261	20	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	261	21	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	261	22	said	say	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	261	23	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	261	24	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	261	25	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	261	26	says	say	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	261	27	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	261	28	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	262	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	262	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	262	3	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	262	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	262	5	score	score	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	6	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	262	7	indicate	indicate	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	262	8	if	if	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	262	10	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	11	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	262	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	262	13	account	account	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	14	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	15	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	262	16	interview	interview	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	262	17	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	263	1	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	263	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	263	3	easy	easy	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	263	4	clues	clue	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	263	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	6	NYT	NYT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	263	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	8	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	263	9	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	10	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	11	paragraphs	paragraph	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	263	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	263	14	inter-	inter-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	263	15	view	view	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	16	portion	portion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	263	19	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	20	beginning	begin	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	263	21	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	22	either	either	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	263	23	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	263	24	Q.	q.	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	25	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	263	26	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	263	27	question	question	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	28	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	263	29	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	263	30	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	263	31	A.	a.	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	32	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	263	33	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	263	34	answer	answer	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	263	35	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	263	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	264	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	2	count	count	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	264	4	total	total	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	5	num-	num-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	6	ber	ber	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	8	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	264	9	Q.	Q.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	10	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	264	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	264	12	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	264	13	A.	a.	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	264	14	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	264	15	prefixes	prefix	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	264	16	combined	combine	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	17	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	264	18	divide	divide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	264	20	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	21	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	264	23	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	264	24	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	264	26	sentences	sentence	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	264	27	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	264	28	INTER-	inter-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	264	29	VIEW	VIEW	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	264	30	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	264	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	265	1	When	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	265	2	either	either	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	265	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	265	4	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	265	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	265	6	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	265	7	Q.	Q.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	265	8	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	265	9	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	265	10	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	265	11	zero	zero	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	265	12	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	265	13	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	265	14	A.	a.	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	265	15	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	265	16	tags	tag	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	265	17	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	265	18	zero	zero	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	265	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	265	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	265	21	score	score	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	265	22	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	265	23	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	265	24	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	265	25	zero	zero	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	265	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	266	1	All	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	266	2	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	266	3	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	266	4	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	266	5	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	266	6	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	266	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	266	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	266	9	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	266	10	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	266	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	267	1	3.5	3.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	267	2	Affective	affective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	267	3	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	4	Some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	267	5	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	267	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	267	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	8	example	example	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	9	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	267	10	detailing	detail	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	267	11	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	12	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	13	health	health	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	267	15	crime	crime	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	267	17	ethics	ethic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	267	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	267	19	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	267	20	provoke	provoke	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	267	21	emotional	emotional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	267	22	re-	re-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	267	23	actions	action	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	267	24	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	25	readers	reader	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	267	26	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	27	shown	show	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	28	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	29	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	267	30	snippet	snippet	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	267	31	below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	267	32	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	268	1	Medicine	medicine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	2	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	268	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	268	4	constant	constant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	268	5	trade	trade	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	268	7	off	off	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	268	9	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	268	10	struggle	struggle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	11	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	268	12	cure	cure	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	268	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	268	14	dis-	dis-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	15	ease	ease	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	16	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	17	killing	kill	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	268	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	268	19	patient	patient	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	268	20	first	first	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	268	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	269	1	Chemotherapy	Chemotherapy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	269	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	269	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	4	exam-	exam-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	269	5	ple	ple	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	269	7	involves	involve	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	269	8	purposely	purposely	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	269	9	poisoning	poison	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	269	10	someone	someone	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	11	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	269	12	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	269	13	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	269	15	ex-	ex-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	269	16	pectation	pectation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	17	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	269	19	short	short	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	269	20	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	269	21	term	term	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	22	injury	injury	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	23	will	will	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	269	24	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	269	25	outweighed	outweigh	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	26	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	269	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	269	28	eventual	eventual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	269	29	benefits	benefit	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	269	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	270	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	270	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	270	3	affect	affect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	270	4	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	270	5	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	270	6	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	270	7	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	270	8	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	270	9	lexicons	lexicon	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	270	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	271	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	271	2	MPQA	MPQA	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	3	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	271	4	Wilson	Wilson	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	5	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	6	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	7	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	271	9	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	271	10	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	271	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	271	12	Gen-	Gen-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	13	eral	eral	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	271	14	Inquirer	Inquirer	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	15	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	271	16	Stone	Stone	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	17	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	18	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	19	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	271	21	1966	1966	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	271	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	271	23	give	give	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	271	24	lists	list	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	271	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	271	26	positive	positive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	271	27	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	271	28	negative	negative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	271	29	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	271	30	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	271	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	272	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	272	2	third	third	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	272	3	resource	resource	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	272	4	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	272	5	emotion	emotion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	272	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	272	7	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	272	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	272	9	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	272	10	FrameNet	FrameNet	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	272	11	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	272	12	Baker	Baker	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	272	13	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	272	14	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	272	15	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	272	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	272	17	1998	1998	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	272	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	272	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	273	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	273	2	sizes	size	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	273	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	273	4	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	273	5	lexicon	lexicon	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	273	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	273	7	8,221	8,221	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	273	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	273	9	5,395	5,395	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	273	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	273	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	273	12	653	653	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	273	13	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	273	14	respectively	respectively	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	273	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	274	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	274	4	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	274	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	6	positive	positive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	274	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	8	negative	negative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	274	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	10	polar	polar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	274	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	274	13	emotion	emotion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	14	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	274	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	16	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	274	17	normalized	normalize	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	18	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	274	20	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	274	21	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	22	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	23	con-	con-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	24	tent	tent	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	25	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	274	26	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	27	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	274	28	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	274	29	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	274	30	POS	POS	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	31	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	32	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	33	NEG	NEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	34	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	35	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	36	PO-	PO-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	37	LAR	LAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	38	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	274	40	EMOT	EMOT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	41	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	274	42	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	274	43	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	275	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	275	3	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	275	5	pro-	pro-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	6	portion	portion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	8	emotion	emotion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	275	10	polar	polar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	275	11	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	275	12	taken	take	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	13	together	together	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	275	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	275	15	POLAR	POLAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	16	EMOT	EMOT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	17	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	275	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	275	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	275	21	ratio	ratio	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	22	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	23	count	count	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	24	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	25	positive	positive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	275	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	275	27	negative	negative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	275	28	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	275	29	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	275	30	POS	POS	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	31	BY	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	275	32	NEG	NEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	275	33	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	275	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	276	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	276	2	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	276	3	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	276	4	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	276	5	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	276	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	276	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	276	8	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	276	9	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	276	10	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	276	11	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	12	NEG	NEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	13	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	276	15	POLAR	POLAR	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	16	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	276	18	EMOT	EMOT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	19	POLAR	POLAR	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	276	20	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	276	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	277	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	277	2	TYPICAL	TYPICAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	277	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	277	5	POS	POS	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	6	BY	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	277	7	NEG	NEG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	277	9	EMOT	EMOT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	10	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	11	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	277	12	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	277	13	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	277	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	278	1	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	278	2	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	278	3	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	278	4	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	278	5	more	more	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	278	6	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	278	7	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	278	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	278	9	mostly	mostly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	278	10	skewed	skew	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	278	11	towards	towards	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	278	12	negative	negative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	278	13	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	278	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	279	1	3.6	3.6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	279	2	Amount	Amount	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	279	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	4	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	5	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	6	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	7	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	8	lay	lay	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	9	audience	audience	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	279	11	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	12	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	13	writer	writer	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	14	presents	present	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	279	15	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	279	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	17	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	279	18	relevant	relevant	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	279	19	findings	finding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	279	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	279	21	methods	method	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	279	22	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	23	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	24	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	25	study	study	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	279	27	interleaves	interleave	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	279	28	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	29	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	30	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	31	de-	de-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	32	tails	tail	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	279	33	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	34	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	35	relevance	relevance	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	36	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	37	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	38	finding	finding	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	279	40	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	279	41	in-	in-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	279	42	volved	volve	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	279	43	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	44	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	45	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	46	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	279	47	general	general	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	279	48	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	49	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	50	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	279	51	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	279	52	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	280	1	As	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	2	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	280	3	result	result	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	280	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	280	6	degree	degree	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	7	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	8	explicit	explicit	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	280	9	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	10	descriptions	description	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	280	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	280	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	280	13	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	280	14	varies	vary	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	280	15	considerably	considerably	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	280	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	281	1	To	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	281	2	test	test	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	281	3	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	281	4	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	281	5	aspect	aspect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	6	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	281	7	associated	associate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	8	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	9	qual-	qual-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	281	10	ity	ity	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	281	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	281	12	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	281	13	count	count	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	281	14	references	reference	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	281	15	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	16	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	17	methods	method	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	281	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	281	19	re-	re-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	281	20	searchers	searcher	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	281	21	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	281	23	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	281	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	282	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	282	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	282	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	282	4	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	5	dictio-	dictio-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	6	nary	nary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	282	7	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	8	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	282	9	introduced	introduce	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	282	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	11	Section	section	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	12	2	2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	282	13	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	282	15	source	source	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	17	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	18	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	282	19	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	282	20	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	282	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	283	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	2	count	count	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	283	4	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	283	5	num-	num-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	6	348	348	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	283	7	ber	ber	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	8	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	283	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	283	12	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	13	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	283	14	match	match	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	15	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	283	16	dictionary	dictionary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	283	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	283	18	RES	RES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	19	TOTAL	TOTAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	20	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	283	21	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	283	22	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	283	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	283	24	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	26	unique	unique	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	283	27	match-	match-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	28	ing	ing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	283	29	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	283	30	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	283	31	RES	RES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	32	UNIQ	UNIQ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	283	33	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	283	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	284	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	284	3	normalize	normalize	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	4	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	284	5	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	284	6	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	284	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	284	8	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	284	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	284	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	284	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	284	12	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	284	13	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	284	14	create	create	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	284	15	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	284	16	RES	RES	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	17	TOTAL	TOTAL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	18	PROP	PROP	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	284	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	284	20	RES	res	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	284	21	UNIQ	UNIQ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	22	PROP	PROP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	284	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	285	1	All	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	285	2	four	four	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	285	3	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	285	4	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	285	5	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	285	6	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	285	7	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	285	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	285	10	VERY	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	285	11	GOOD	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	285	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	285	13	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	285	14	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	285	15	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	16	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	285	17	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	285	18	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	285	19	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	285	20	associated	associate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	21	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	22	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	285	23	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	285	24	amount	amount	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	25	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	26	direct	direct	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	285	27	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	28	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	285	29	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	285	30	explanations	explanation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	285	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	286	1	4	4	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	286	2	Classification	classification	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	3	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	4	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	286	5	trained	train	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	286	6	classifiers	classifier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	286	7	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	286	8	all	all	PDT
work_hea3gijzhnc2rl4sxgyj6pvghm	286	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	286	10	above	above	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	286	11	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	286	12	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	13	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	286	15	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	286	16	settings–‘any	settings–‘any	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	17	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	286	18	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	19	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	286	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	286	21	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	286	22	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	286	23	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	286	24	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	25	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	286	26	in-	in-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	27	troduced	troduce	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	286	28	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	286	29	Section	Section	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	286	30	2.3	2.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	286	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	287	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	287	2	baseline	baseline	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	287	3	random	random	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	287	4	accu-	accu-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	287	5	racy	racy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	287	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	287	7	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	287	8	cases	case	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	287	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	287	10	50	50	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	287	11	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	287	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	288	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	2	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	3	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	288	4	SVM	SVM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	5	classi-	classi-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	6	fier	fier	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	7	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	288	9	radial	radial	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	288	10	basis	basis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	11	kernel	kernel	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	12	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	288	13	R	R	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	14	Development	Development	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	15	Core	Core	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	16	Team	Team	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	288	17	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	288	18	2011	2011	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	288	19	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	288	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	288	21	parameters	parameter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	288	22	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	288	23	tuned	tune	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	24	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	288	25	cross	cross	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	26	validation	validation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	27	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	28	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	288	29	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	288	30	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	288	31	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	289	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	289	2	best	good	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	289	3	parameters	parameter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	289	4	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	289	5	then	then	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	289	6	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	7	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	289	8	classify	classify	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	289	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	289	10	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	11	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	289	14	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	289	15	fold	fold	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	289	16	cross	cross	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	17	-	-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	289	18	validation	validation	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	289	19	setting	setting	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	289	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	290	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	290	2	di-	di-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	290	3	vide	vide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	290	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	290	5	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	6	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	7	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	8	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	290	9	parts	part	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	290	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	290	11	train	train	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	290	12	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	13	9	9	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	290	14	parts	part	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	290	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	290	16	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	17	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	290	19	held	hold	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	290	20	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	290	21	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	290	22	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	290	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	291	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	291	2	average	average	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	291	3	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	291	4	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	291	6	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	291	7	experiments	experiment	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	291	8	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	291	9	75.3	75.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	291	10	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	11	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	12	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	291	14	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	291	15	any-	any-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	291	16	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	17	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	291	18	setup	setup	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	291	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	291	21	68	68	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	291	22	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	23	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	24	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	25	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	291	26	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	27	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	291	28	paired	pair	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	29	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	291	30	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	291	31	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	291	32	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	33	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	291	34	setup	setup	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	291	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	292	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	292	2	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	292	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	292	4	consider-	consider-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	292	5	able	able	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	292	6	improvements	improvement	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	292	7	over	over	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	292	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	292	9	baseline	baseline	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	292	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	293	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	293	2	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	293	3	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	293	4	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	293	5	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	293	6	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	293	7	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	293	8	contains	contain	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	293	9	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	293	10	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	293	11	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	293	12	varying	varying	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	293	13	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	293	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	294	1	So	so	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	294	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	294	3	investigate	investigate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	294	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	294	5	relationship	relationship	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	6	between	between	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	7	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	8	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	294	10	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	12	prediction	prediction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	13	more	more	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	294	14	closely	closely	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	294	15	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	16	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	294	17	setting	setting	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	294	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	295	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	295	2	divide	divide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	295	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	295	4	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	5	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	295	6	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	7	bins	bin	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	295	8	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	9	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	295	11	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	12	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	295	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	296	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	296	2	compute	compute	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	296	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	296	4	10-fold	10-fold	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	296	5	cross	cross	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	296	6	validation	validation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	296	7	predictions	prediction	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	296	8	us-	us-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	296	9	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	296	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	296	11	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	296	12	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	296	13	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	296	14	above	above	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	296	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	296	16	collect	collect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	296	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	296	18	predicted	predict	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	296	19	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	296	20	across	across	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	296	21	all	all	PDT
work_hea3gijzhnc2rl4sxgyj6pvghm	296	22	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	296	23	folds	fold	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	296	24	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	297	1	Then	then	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	297	2	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	297	3	com-	com-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	297	4	pute	pute	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	297	5	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	297	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	297	7	examples	example	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	297	8	within	within	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	297	9	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	297	10	bin	bin	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	297	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	298	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	298	2	results	result	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	298	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	298	4	plotted	plot	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	6	Figure	Figure	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	298	7	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	298	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	298	9	int	int	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	10	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	298	11	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	12	refers	refer	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	298	13	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	298	15	full	full	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	298	16	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	17	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	18	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	298	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	298	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	298	21	results	result	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	298	22	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	298	24	six	six	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	298	25	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	26	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	298	27	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	298	28	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	298	29	indicated	indicate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	298	30	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	299	1	As	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	299	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	299	3	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	299	4	increases	increase	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	299	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	299	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	299	7	prediction	prediction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	299	8	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	299	9	be-	be-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	299	10	comes	come	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	299	11	harder	hard	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	299	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	300	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	300	2	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	4	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	300	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	300	6	gives	give	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	300	7	66	66	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	300	8	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	9	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	11	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	300	12	above	above	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	13	0.4	0.4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	300	14	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	300	16	74	74	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	300	17	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	18	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	300	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	300	20	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	21	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	300	22	less	less	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	300	23	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	300	24	0.15	0.15	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	300	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	301	1	Most	Most	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	301	2	individ-	individ-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	301	3	ual	ual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	301	4	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	301	5	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	301	6	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	301	7	show	show	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	301	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	301	9	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	301	10	trend	trend	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	301	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	302	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	302	2	result	result	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	302	3	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	302	4	understandable	understandable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	302	5	because	because	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	302	6	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	302	7	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	302	8	simi-	simi-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	302	9	lar	lar	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	302	10	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	302	11	could	could	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	302	12	exhibit	exhibit	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	302	13	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	302	14	properties	property	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	302	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	303	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	2	ex-	ex-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	303	3	ample	ample	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	303	5	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	303	6	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	7	about	about	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	8	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	303	9	controversies	controversy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	10	surround-	surround-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	303	11	ing	ing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	303	12	vaccination	vaccination	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	303	13	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	303	14	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	303	15	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	16	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	303	17	have	have	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	303	18	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	19	levels	level	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	20	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	21	people	people	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	22	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	303	23	oriented	orient	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	24	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	25	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	303	26	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	27	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	28	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	303	29	narrative	narrative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	30	style	style	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	31	Figure	figure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	32	1	1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	303	33	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	303	34	Accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	35	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	36	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	37	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	38	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	39	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	40	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	41	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	303	42	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	43	way	way	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	44	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	45	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	303	46	space	space	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	47	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	303	48	related	relate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	49	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	303	50	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	303	51	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	303	52	likely	likely	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	53	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	303	54	contain	contain	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	303	55	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	56	visual	visual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	303	57	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	303	58	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	304	1	There	there	EX
work_hea3gijzhnc2rl4sxgyj6pvghm	304	2	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	304	3	however	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	304	4	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	304	5	exceptions	exception	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	304	6	—	—	:
work_hea3gijzhnc2rl4sxgyj6pvghm	304	7	affect	affect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	304	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	304	9	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	304	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	305	1	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	2	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	305	3	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	305	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	305	6	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	7	improve	improve	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	305	8	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	9	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	305	10	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	11	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	305	12	affect	affect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	305	13	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	14	give	give	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	305	15	51	51	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	16	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	18	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	19	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	20	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	21	0.1	0.1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	22	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	305	23	62	62	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	24	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	25	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	26	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	27	above	above	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	28	0.4	0.4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	29	similar-	similar-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	30	ity	ity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	305	32	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	33	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	34	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	35	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	36	goes	go	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	305	37	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	38	52	52	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	39	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	40	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	41	57	57	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	305	42	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	43	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	44	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	305	45	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	305	46	similarity	similarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	305	47	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	305	48	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	306	1	This	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	306	2	finding	find	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	306	3	il-	il-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	306	4	lustrates	lustrate	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	306	5	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	6	even	even	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	306	7	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	306	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	9	very	very	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	306	10	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	306	11	topics	topic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	306	12	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	306	13	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	306	14	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	15	differently	differently	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	306	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	306	17	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	18	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	306	19	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	306	20	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	21	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	306	22	excel-	excel-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	23	lent	lend	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	306	24	authors	author	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	306	25	associated	associate	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	26	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	27	greater	great	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	306	28	degree	degree	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	29	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	30	senti-	senti-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	306	31	ment	ment	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	306	32	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	306	33	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	306	34	deeper	deep	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	306	35	study	study	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	36	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	37	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	306	38	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	39	problem	problem	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	306	40	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	307	1	5	5	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	307	2	Combining	combine	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	307	3	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	307	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	5	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	6	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	7	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	307	8	now	now	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	307	9	compare	compare	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	307	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	307	11	combine	combine	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	307	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	307	13	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	14	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	307	15	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	307	16	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	17	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	307	18	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	19	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	307	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	307	21	41	41	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	307	22	total	total	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	23	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	307	24	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	25	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	307	26	dis-	dis-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	27	cussed	cuss	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	28	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	29	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	30	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	31	readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	32	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	307	33	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	307	34	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	307	35	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	36	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	37	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	307	38	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	39	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	307	40	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	41	classification	classification	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	307	42	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	308	1	Readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	2	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	308	3	16	16	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	308	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	308	5	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	308	6	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	308	7	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	308	8	test	test	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	308	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	308	10	full	full	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	308	11	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	13	readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	14	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	308	15	studied	study	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	16	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	17	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	308	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	308	19	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	308	20	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	308	21	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	308	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	308	23	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	308	24	involving	involve	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	308	25	token	token	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	26	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	308	27	type	type	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	28	ratio	ratio	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	29	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	308	30	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	31	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	308	32	sen-	sen-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	33	tence	tence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	34	length	length	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	35	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	308	36	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	37	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	38	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	308	39	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	308	40	cohesion	cohesion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	41	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	308	42	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	308	43	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	308	44	estimates	estimate	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	308	45	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	46	complexity	complexity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	308	47	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	309	1	Well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	309	2	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	309	3	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	4	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	5	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	309	6	23	23	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	309	7	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	309	8	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	309	9	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	309	10	For	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	11	well-	well-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	309	12	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	13	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	309	15	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	309	16	use	use	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	309	17	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	309	18	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	309	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	20	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	309	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	309	22	both	both	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	309	23	related	related	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	309	24	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	25	discourse	discourse	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	309	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	310	1	One	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	310	2	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	310	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	310	4	probabilities	probability	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	310	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	6	dif-	dif-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	310	7	ferent	ferent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	310	8	types	type	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	310	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	10	entity	entity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	11	transitions	transition	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	310	12	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	310	14	Entity	Entity	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	15	Grid	Grid	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	16	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	17	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	310	18	Barzilay	Barzilay	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	310	20	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	310	22	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	310	23	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	310	24	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	310	25	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	26	com-	com-	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	27	pute	pute	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	28	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	310	29	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	310	30	Brown	Brown	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	31	Coherence	Coherence	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	32	Toolkit	Toolkit	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	33	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	310	34	Elsner	Elsner	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	35	et	et	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	36	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	37	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	310	38	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	310	39	2007	2007	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	310	40	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	310	41	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	311	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	311	2	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	311	3	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	311	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	311	7	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	311	8	de-	de-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	311	9	fined	fine	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	11	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	311	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	311	13	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	311	14	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	311	15	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	311	16	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	311	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	18	likelihoods	likelihood	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	311	19	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	311	20	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	311	21	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	311	22	explicit	explicit	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	311	23	discourse	discourse	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	311	24	relations	relation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	311	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	312	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	312	2	iden-	iden-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	312	3	tified	tifie	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	312	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	312	5	relations	relation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	312	6	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	312	7	texts	text	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	312	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	312	9	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	312	10	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	312	11	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	312	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	312	13	349	349	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	312	14	AddDiscourse	AddDiscourse	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	312	15	tool	tool	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	312	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	312	17	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	312	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	312	19	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	312	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	312	21	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	312	22	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	312	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	313	1	Interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	313	2	fiction	fiction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	3	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	313	4	22	22	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	313	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	313	6	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	313	7	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	313	8	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	313	9	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	313	10	intro-	intro-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	11	duced	duce	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	12	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	13	McIntyre	McIntyre	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	313	14	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	313	15	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	313	16	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	313	17	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	313	18	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	313	19	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	20	predicting	predict	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	313	21	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	22	ratings	rating	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	313	23	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	24	fairy	fairy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	313	25	tales	tale	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	313	26	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	314	1	They	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	314	2	include	include	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	314	3	counts	count	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	314	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	314	5	syntactic	syntactic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	314	6	items	item	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	314	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	314	8	relations	relation	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	314	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	314	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	314	11	token	token	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	314	12	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	314	13	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	314	14	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	314	15	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	314	16	psycholinguistic	psycholinguistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	314	17	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	314	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	315	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	315	2	nor-	nor-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	315	3	malize	malize	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	315	4	each	each	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	315	5	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	315	6	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	315	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	315	8	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	315	9	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	315	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	315	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	315	12	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	315	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	316	1	Content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	316	2	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	316	3	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	316	4	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	316	5	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	316	6	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	316	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	316	8	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	316	9	present	present	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	316	10	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	316	11	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	316	12	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	316	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	317	1	Word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	2	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	317	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	4	standard	standard	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	317	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	6	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	317	8	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	9	recommendation	recommendation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	10	systems	system	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	317	11	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	317	12	Pazzani	Pazzani	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	13	et	et	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	317	14	al	al	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	15	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	317	17	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	317	18	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	317	19	Mooney	Mooney	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	20	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	317	21	Roy	Roy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	317	23	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	317	24	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	317	25	where	where	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	317	26	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	27	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	28	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	29	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	317	30	pick	pick	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	317	31	out	out	RP
work_hea3gijzhnc2rl4sxgyj6pvghm	317	32	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	317	33	similar	similar	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	317	34	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	35	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	317	36	which	which	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	317	37	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	317	38	user	user	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	39	has	have	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	317	40	already	already	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	317	41	read	read	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	317	42	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	318	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	318	2	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	318	3	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	318	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	318	5	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	318	6	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	318	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	318	8	most	most	RBS
work_hea3gijzhnc2rl4sxgyj6pvghm	318	9	frequent	frequent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	318	10	n	n	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	318	11	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	318	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	318	13	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	318	14	corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	318	15	after	after	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	318	16	removing	remove	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	318	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	318	18	50	50	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	318	19	most	most	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	318	20	frequent	frequent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	318	21	ones	one	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	318	22	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	319	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	319	2	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	3	’s	’s	POS
work_hea3gijzhnc2rl4sxgyj6pvghm	319	4	count	count	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	6	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	319	7	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	8	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	319	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	319	10	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	11	value	value	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	319	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	320	1	Note	note	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	320	2	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	3	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	4	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	320	5	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	320	6	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	7	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	320	8	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	320	9	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	10	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	320	11	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	320	14	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	15	such	such	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	320	16	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	17	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	18	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	320	19	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	320	20	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	321	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	321	2	random	random	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	321	3	sample	sample	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	5	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	321	6	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	7	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	321	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	9	n	n	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	10	=	=	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	321	11	1000	1000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	321	12	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	321	13	shown	show	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	14	below	below	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	321	15	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	321	16	reflects	reflect	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	321	17	this	this	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	321	18	aspect	aspect	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	321	19	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	322	1	“	"	``
work_hea3gijzhnc2rl4sxgyj6pvghm	322	2	matter	matter	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	322	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	4	series	series	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	6	wear	wear	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	8	nation	nation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	10	ac-	ac-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	322	11	count	count	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	12	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	13	surgery	surgery	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	14	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	15	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	322	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	17	receive	receive	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	322	18	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	19	remember	remember	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	322	20	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	21	support	support	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	23	worry	worry	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	24	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	25	enough	enough	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	322	26	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	27	office	office	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	29	prevent	prevent	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	322	30	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	31	biggest	big	JJS
work_hea3gijzhnc2rl4sxgyj6pvghm	322	32	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	322	33	customer	customer	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	322	34	”	"	''
work_hea3gijzhnc2rl4sxgyj6pvghm	322	35	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	323	1	Table	table	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	2	2	2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	323	3	compares	compare	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	323	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	323	5	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	7	SVM	SVM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	323	8	classi-	classi-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	323	9	fiers	fier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	10	trained	train	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	323	11	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	12	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	13	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	14	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	323	15	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	16	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	323	17	their	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	323	18	combinations.6	combinations.6	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	19	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	323	20	readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	21	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	323	22	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	323	23	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	323	24	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	25	nature	nature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	26	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	323	27	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	323	28	fiction	fiction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	29	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	30	provide	provide	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	323	31	good	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	323	32	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	323	33	60	60	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	323	34	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	323	35	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	323	36	above	above	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	323	37	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	324	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	324	2	genre	genre	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	324	3	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	324	4	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	324	5	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	324	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	324	7	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	324	8	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	324	9	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	324	10	individually	individually	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	324	11	much	much	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	324	12	stronger	strong	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	324	13	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	324	14	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	324	15	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	324	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	325	1	Different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	325	2	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	3	as-	as-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	325	4	pects	pect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	325	5	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	325	6	without	without	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	7	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	8	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	325	9	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	325	10	clearly	clearly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	325	11	complementary	complementary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	325	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	325	13	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	325	14	combined	combine	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	15	give	give	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	325	16	76	76	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	325	17	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	18	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	19	79	79	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	325	20	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	21	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	22	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	23	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	325	24	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	325	25	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	325	26	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	325	27	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	28	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	325	29	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	30	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	325	31	74	74	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	325	32	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	33	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	34	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	325	35	topic	topic	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	325	36	pairs	pair	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	325	37	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	325	38	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	326	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	326	2	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	326	3	bag	bag	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	5	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	326	6	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	326	7	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	8	remark-	remark-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	326	9	ably	ably	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	326	10	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	326	11	giving	give	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	326	12	80	80	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	326	13	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	14	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	15	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	326	16	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	326	17	settings	setting	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	326	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	327	1	As	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	2	mentioned	mention	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	3	before	before	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	4	these	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	327	5	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	6	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	327	7	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	327	9	mix	mix	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	11	topic	topic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	327	12	indicators	indicator	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	327	13	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	327	14	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	327	15	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	16	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	327	17	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	18	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	327	20	ar-	ar-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	21	ticles	ticle	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	327	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	327	23	ie	ie	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	24	.	.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	327	26	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	27	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	327	28	implicitly	implicitly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	327	29	indicate	indicate	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	327	30	animacy	animacy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	327	32	re-	re-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	327	33	search	search	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	34	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	327	35	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	327	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	328	1	But	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	328	2	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	328	3	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	328	4	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	6	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	7	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	328	8	above	above	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	9	all	all	PDT
work_hea3gijzhnc2rl4sxgyj6pvghm	328	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	328	11	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	12	categories	category	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	328	13	indicates	indicate	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	328	14	that	that	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	15	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	16	plays	play	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	328	17	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	328	18	important	important	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	328	19	role	role	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	20	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	21	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	22	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	328	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	329	1	However	however	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	329	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	329	3	despite	despite	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	329	5	high	high	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	329	6	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	329	8	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	9	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	329	10	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	329	11	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	329	12	easily	easily	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	329	13	interpretable	interpretable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	329	14	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	15	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	329	16	classes	class	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	329	17	re-	re-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	329	18	lated	late	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	329	19	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	20	writing	write	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	329	21	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	22	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	329	23	have	have	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	329	24	done	do	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	25	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	26	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	329	27	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	329	28	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	329	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	330	1	Further	further	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	330	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	330	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	330	4	total	total	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	330	5	set	set	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	7	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	8	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	9	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	330	10	6For	6For	,
work_hea3gijzhnc2rl4sxgyj6pvghm	330	11	classifiers	classifier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	12	involving	involve	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	330	13	content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	14	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	15	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	330	16	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	330	17	did	do	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	330	18	not	not	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	330	19	tune	tune	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	330	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	330	21	SVM	SVM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	330	22	parameters	parameter	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	23	because	because	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	24	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	25	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	330	26	small	small	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	330	27	size	size	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	28	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	29	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	30	data	datum	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	31	compared	compare	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	32	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	33	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	34	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	330	35	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	330	36	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	331	1	Default	Default	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	331	2	SVM	SVM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	331	3	settings	setting	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	331	4	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	331	5	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	331	6	instead	instead	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	331	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	332	1	Feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	2	set	set	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	3	Any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	332	4	Topic	topic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	5	Same	Same	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	6	Interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	8	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	9	75.3	75.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	10	68.0	68.0	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	11	Readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	12	65.5	65.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	13	63.0	63.0	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	14	Well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	332	15	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	16	written	write	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	17	59.1	59.1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	18	59.9	59.9	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	19	Interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	20	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	21	fiction	fiction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	22	67.9	67.9	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	23	62.8	62.8	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	24	Readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	25	+	+	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	332	26	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	332	27	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	28	writ	writ	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	29	64.7	64.7	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	30	64.3	64.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	31	Readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	32	+	+	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	332	33	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	332	34	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	35	writ	writ	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	36	+	+	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	332	37	Int	int	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	38	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	39	fict	fict	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	40	71.0	71.0	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	41	70.3	70.3	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	42	Readable	readable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	43	+	+	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	332	44	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	332	45	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	46	writ	writ	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	47	+	+	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	332	48	Int	Int	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	49	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	332	50	sci	sci	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	51	79.5	79.5	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	52	73.2	73.2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	53	All	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	332	54	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	55	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	56	76.7	76.7	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	57	74.7	74.7	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	58	Content	Content	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	59	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	60	500	500	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	61	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	62	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	63	81.7	81.7	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	64	79.4	79.4	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	65	Content	Content	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	66	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	67	1000	1000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	68	words	word	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	69	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	70	81.2	81.2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	71	82.1	82.1	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	72	Combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	73	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	332	74	Writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	75	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	76	all	all	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	332	77	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	78	+	+	NFP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	79	Content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	80	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	81	1000w	1000w	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	82	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	332	83	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	84	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	85	vector	vector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	86	82.6	82.6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	87	*	*	NFP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	88	84.0	84.0	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	89	*	*	NFP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	90	Sum	sum	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	91	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	92	confidence	confidence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	93	scores	score	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	94	81.6	81.6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	95	84.9	84.9	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	96	Oracle	Oracle	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	332	97	87.6	87.6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	98	93.8	93.8	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	99	Table	table	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	100	2	2	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	101	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	332	102	Accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	103	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	104	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	332	105	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	106	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	107	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	108	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	332	109	102	102	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	110	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	111	contrast	contrast	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	112	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	113	1000	1000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	332	114	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	332	115	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	332	116	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	333	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	2	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	333	3	interest	interest	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	4	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	333	5	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	6	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	7	set	set	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	333	9	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	333	10	aimed	aim	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	333	11	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	333	12	highlight	highlight	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	333	13	how	how	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	333	14	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	333	15	some	some	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	333	16	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	17	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	333	18	aspects	aspect	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	333	19	considered	consider	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	20	important	important	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	333	21	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	22	good	good	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	333	23	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	24	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	25	can	can	MD
work_hea3gijzhnc2rl4sxgyj6pvghm	333	26	predict	predict	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	333	27	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	333	28	ratings	rating	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	333	29	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	334	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	334	2	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	334	3	combined	combine	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	334	4	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	334	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	334	6	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	334	7	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	334	8	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	334	9	mix	mix	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	334	10	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	334	11	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	334	12	writing	write	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	334	13	related	related	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	334	14	predictors	predictor	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	334	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	335	1	We	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	2	do	do	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	3	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	4	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	6	three	three	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	335	7	ways	way	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	8	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	9	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	10	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	335	12	writing	write	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	335	13	fea-	fea-	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	14	tures	ture	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	15	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	16	included	include	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	17	together	together	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	335	18	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	19	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	20	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	21	vector	vector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	22	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	335	23	b	b	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	24	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	25	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	335	26	separate	separate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	335	27	classifiers	classifier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	28	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	29	trained	train	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	30	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	31	one	one	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	335	32	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	335	33	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	34	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	35	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	335	36	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	37	other	other	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	335	38	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	335	39	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	40	ones	one	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	41	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	42	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	335	43	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	44	sum	sum	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	45	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	46	confidence	confidence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	47	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	48	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	335	49	used	use	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	50	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	335	51	decide	decide	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	335	52	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	53	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	54	final	final	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	335	55	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	56	;	;	:
work_hea3gijzhnc2rl4sxgyj6pvghm	335	57	c	c	LS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	58	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	59	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	60	oracle	oracle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	61	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	62	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	335	63	two	two	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	335	64	classifiers	classifier	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	335	65	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	66	trained	train	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	67	just	just	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	335	68	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	69	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	70	option	option	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	71	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	72	b	b	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	73	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	335	74	but	but	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	335	75	when	when	WRB
work_hea3gijzhnc2rl4sxgyj6pvghm	335	76	they	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	77	disagree	disagree	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	78	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	79	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	80	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	81	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	335	82	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	83	pick	pick	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	335	84	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	335	85	correct	correct	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	335	86	label	label	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	335	87	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	336	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	336	2	oracle	oracle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	3	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	4	gives	give	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	336	5	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	336	6	simple	simple	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	336	7	upper	upper	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	336	8	bound	bind	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	9	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	10	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	336	11	accuracy	accuracy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	12	obtainable	obtainable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	336	13	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	14	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	336	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	337	1	These	these	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	337	2	values	value	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	337	3	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	337	4	87	87	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	337	5	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	6	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	7	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	337	8	any	any	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	337	9	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	337	10	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	11	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	337	12	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	337	13	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	337	14	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	337	15	93.8	93.8	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	337	16	%	%	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	18	‘	'	``
work_hea3gijzhnc2rl4sxgyj6pvghm	337	19	same	same	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	337	20	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	337	21	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	337	22	’	'	''
work_hea3gijzhnc2rl4sxgyj6pvghm	337	23	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	338	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	338	2	automatic	automatic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	338	3	methods	method	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	338	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	338	5	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	338	6	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	7	vector	vector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	8	combi-	combi-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	9	nation	nation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	338	11	classifier	classifier	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	12	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	13	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	338	14	give	give	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	338	15	better	well	RBR
work_hea3gijzhnc2rl4sxgyj6pvghm	338	16	ac-	ac-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	338	17	curacies	curacy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	338	18	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	19	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	338	20	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	338	21	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	22	or	or	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	338	23	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	338	24	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	338	25	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	339	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	339	2	accuracies	accuracy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	339	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	339	5	folds	fold	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	339	6	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	7	10	10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	339	8	fold	fold	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	339	9	cross	cross	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	10	valida-	valida-	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	339	11	tion	tion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	12	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	13	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	339	14	feature	feature	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	15	vector	vector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	16	combination	combination	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	17	method	method	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	18	were	be	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	339	19	also	also	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	339	20	found	find	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	21	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	339	22	be	be	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	339	23	significantly	significantly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	339	24	higher	high	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	339	25	than	than	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	26	those	those	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	339	27	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	28	word	word	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	29	features	feature	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	339	30	only	only	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	339	31	(	(	-LRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	339	32	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	339	33	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	339	34	paired	pair	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	35	Wilcoxon	Wilcoxon	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	339	36	signed-	signed-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	37	rank	rank	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	38	test	test	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	339	39	)	)	-RRB-
work_hea3gijzhnc2rl4sxgyj6pvghm	339	40	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	340	1	Therefore	therefore	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	340	2	both	both	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	340	3	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	340	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	340	5	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	340	6	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	340	7	are	be	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	340	8	clearly	clearly	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	340	9	useful	useful	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	340	10	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	340	11	identifying	identify	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	340	12	great	great	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	340	13	articles	article	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	340	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	341	1	6	6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	341	2	Conclusion	Conclusion	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	341	3	Our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	341	4	work	work	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	5	is	be	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	341	6	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	341	7	step	step	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	8	towards	towards	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	9	measuring	measure	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	341	10	overall	overall	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	341	11	arti-	arti-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	12	cle	cle	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	13	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	14	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	15	showing	show	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	341	16	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	341	17	complementary	complementary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	341	18	benefits	benefit	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	341	19	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	20	general	general	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	341	21	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	341	22	domain	domain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	23	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	341	24	specific	specific	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	341	25	writing	write	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	341	26	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	341	27	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	341	28	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	341	29	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	30	indicators	indicator	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	341	31	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	32	article	article	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	33	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	341	34	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	342	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	2	future	future	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	3	we	-PRON-	PRP
work_hea3gijzhnc2rl4sxgyj6pvghm	342	4	plan	plan	VBP
work_hea3gijzhnc2rl4sxgyj6pvghm	342	5	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	342	6	focus	focus	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	342	7	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	8	development	development	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	10	more	more	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	342	11	features	feature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	342	12	as	as	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	342	13	well	well	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	342	14	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	15	better	well	JJR
work_hea3gijzhnc2rl4sxgyj6pvghm	342	16	methods	method	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	342	17	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	342	18	combining	combine	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	342	19	different	different	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	342	20	measures	measure	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	342	21	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	343	1	350	350	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	343	2	References	References	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	343	3	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	4	F.	F.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	5	Baker	Baker	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	343	7	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	8	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	9	Fillmore	Fillmore	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	343	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	343	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	343	12	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	344	1	B.	B.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	344	2	Lowe	Lowe	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	344	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	345	1	1998	1998	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	345	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	346	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	346	2	berkeley	berkeley	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	346	3	framenet	framenet	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	346	4	project	project	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	346	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	347	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	347	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	347	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	347	4	COLING	COLING	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	347	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	347	6	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	347	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	347	8	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	347	9	86–90	86–90	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	347	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	348	1	Z.	Z.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	2	Bao	Bao	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	348	4	B.	B.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	5	Kimelfeld	Kimelfeld	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	348	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	348	8	Y.	Y.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	9	Li	Li	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	348	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	349	1	2011	2011	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	349	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	350	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	350	2	graph	graph	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	3	ap-	ap-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	350	4	proach	proach	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	5	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	6	spelling	spell	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	350	7	correction	correction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	8	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	9	domain	domain	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	10	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	350	11	centric	centric	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	12	search	search	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	350	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	351	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	351	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	351	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	351	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	351	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	351	6	HLT	HLT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	351	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	351	8	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	351	9	905–914	905–914	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	351	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	352	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	352	2	Barzilay	Barzilay	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	352	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	352	4	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	352	5	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	352	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	353	1	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	353	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	354	1	Modeling	model	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	354	2	local	local	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	354	3	coher-	coher-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	354	4	ence	ence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	354	5	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	354	6	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	354	7	entity	entity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	354	8	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	354	9	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	354	10	approach	approach	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	354	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	355	1	Computational	computational	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	355	2	Lin-	Lin-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	355	3	guistics	guistic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	355	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	355	5	34(1):1–34	34(1):1–34	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	355	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	356	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	2	Barzilay	Barzilay	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	356	4	N.	N.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	5	Elhadad	Elhadad	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	356	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	356	8	K.	K.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	9	McKeown	McKeown	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	356	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	357	1	2002	2002	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	357	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	358	1	Inferring	infer	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	358	2	strategies	strategy	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	358	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	4	sentence	sentence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	5	ordering	order	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	358	6	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	7	multi-	multi-	FW
work_hea3gijzhnc2rl4sxgyj6pvghm	358	8	document	document	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	9	summar	summar	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	10	ization	ization	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	358	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	359	1	Journal	Journal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	359	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	359	3	Artificial	Artificial	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	359	4	Intel-	Intel-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	359	5	ligence	ligence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	359	6	Research	Research	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	359	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	359	8	17:35–55	17:35–55	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	359	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	360	1	D.M.	D.M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	361	1	Blei	Blei	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	361	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	361	3	A.Y.	A.Y.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	362	1	Ng	ng	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	362	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	362	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	362	4	M.I.	M.I.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	363	1	Jordan	Jordan	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	363	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	364	1	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	364	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	365	1	Latent	latent	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	365	2	dirichlet	dirichlet	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	365	3	allocation	allocation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	365	4	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	366	1	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	366	2	Journal	Journal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	366	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	366	4	machine	machine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	366	5	Learning	Learning	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	366	6	research	research	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	366	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	366	8	3:993–1022	3:993–1022	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	366	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	367	1	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	2	Blum	Blum	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	367	4	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	5	Knudson	Knudson	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	367	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	367	8	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	9	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	10	Henig	Henig	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	367	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	367	12	editors	editor	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	367	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	368	1	2006	2006	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	368	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	369	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	369	2	field	field	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	3	guide	guide	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	4	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	5	science	science	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	6	writers	writer	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	369	7	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	369	8	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	369	9	official	official	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	369	10	guide	guide	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	11	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	12	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	369	13	national	national	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	369	14	association	association	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	369	15	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	369	16	science	science	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	369	17	writers	writer	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	369	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	370	1	Oxford	Oxford	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	370	2	University	University	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	370	3	Press	Press	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	370	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	370	5	New	New	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	370	6	York	York	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	370	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	371	1	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	371	2	Brill	Brill	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	371	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	371	4	R.C.	R.C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	372	1	Moore	Moore	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	372	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	373	1	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	373	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	374	1	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	374	2	improved	improved	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	374	3	error	error	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	374	4	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	374	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	374	6	noisy	noisy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	374	7	channel	channel	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	374	8	spelling	spelling	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	374	9	correction	correction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	374	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	375	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	375	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	375	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	375	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	375	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	375	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	375	7	286–293	286–293	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	375	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	376	1	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	376	2	F.	F.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	376	3	Chen	Chen	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	376	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	376	5	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	376	6	Goodman	Goodman	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	376	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	377	1	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	377	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	378	1	An	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	378	2	empirical	empirical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	378	3	study	study	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	378	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	378	5	smoothing	smooth	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	378	6	techniques	technique	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	378	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	378	8	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	378	9	modeling	modeling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	378	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	379	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	379	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	379	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	379	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	379	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	379	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	379	7	310–318	310–318	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	379	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	380	1	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	380	2	Cucerzan	Cucerzan	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	380	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	380	4	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	380	5	Brill	Brill	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	380	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	381	1	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	381	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	382	1	Spelling	spell	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	382	2	correction	correction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	3	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	4	an	an	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	382	5	iterative	iterative	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	382	6	process	process	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	7	that	that	WDT
work_hea3gijzhnc2rl4sxgyj6pvghm	382	8	exploits	exploit	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	382	9	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	382	10	collective	collective	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	382	11	knowledge	knowledge	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	12	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	13	web	web	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	382	14	users	user	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	382	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	383	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	383	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	383	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	383	4	EMNLP	EMNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	383	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	383	6	pages	page	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	383	7	293	293	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	383	8	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	383	9	300	300	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	383	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	384	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	384	2	Dale	Dale	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	384	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	384	4	A.	A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	384	5	Kilgarriff	Kilgarriff	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	384	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	385	1	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	385	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	386	1	Helping	help	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	386	2	our	-PRON-	PRP$
work_hea3gijzhnc2rl4sxgyj6pvghm	386	3	own	own	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	386	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	386	5	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	386	6	massaging	massaging	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	386	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	386	8	computational	computational	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	386	9	linguistics	linguistic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	386	10	as	as	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	386	11	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	386	12	new	new	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	386	13	shared	shared	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	386	14	task	task	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	386	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	387	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	387	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	387	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	387	4	INLG	INLG	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	387	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	387	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	387	7	263–267	263–267	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	387	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	388	1	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	2	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	3	De	De	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	4	Marneffe	Marneffe	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	388	6	B.	B.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	7	MacCartney	MacCartney	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	388	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	388	10	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	11	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	12	Man-	Man-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	388	13	ning	ning	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	388	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	389	1	2006	2006	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	389	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	390	1	Generating	generate	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	390	2	typed	type	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	390	3	dependency	dependency	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	390	4	parses	parse	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	390	5	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	390	6	phrase	phrase	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	390	7	structure	structure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	390	8	parses	parse	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	390	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	391	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	391	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	391	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	391	4	LREC	LREC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	391	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	391	6	vol-	vol-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	391	7	ume	ume	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	391	8	6	6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	391	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	391	10	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	391	11	449–454	449–454	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	391	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	392	1	P.	P.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	392	2	Diederich	Diederich	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	392	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	393	1	1974	1974	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	393	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	394	1	Measuring	measure	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	394	2	Growth	Growth	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	394	3	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	394	4	English	English	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	394	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	395	1	Na-	na-	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	395	2	tional	tional	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	395	3	Council	Council	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	395	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	395	5	Teachers	Teachers	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	395	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	395	7	English	English	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	395	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	396	1	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	2	Elsner	Elsner	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	396	4	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	5	Austerweil	Austerweil	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	396	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	396	8	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	9	Charniak	Charniak	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	396	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	397	1	2007	2007	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	397	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	398	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	398	2	uni-	uni-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	398	3	fied	fie	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	398	4	local	local	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	398	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	398	6	global	global	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	398	7	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	398	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	398	9	discourse	discourse	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	398	10	coherence	coherence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	398	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	399	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	399	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	399	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	399	4	NAACL	NAACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	399	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	399	6	HLT	HLT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	399	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	399	8	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	399	9	436–443	436–443	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	399	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	400	1	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	2	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	3	Finkel	Finkel	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	400	5	T.	T.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	6	Grenager	Grenager	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	400	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	400	9	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	10	Manning	Manning	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	400	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	401	1	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	401	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	402	1	In-	in-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	402	2	corporating	corporating	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	402	3	non	non	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	402	4	-	-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	402	5	local	local	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	402	6	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	7	into	into	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	8	information	information	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	9	ex-	ex-	XX
work_hea3gijzhnc2rl4sxgyj6pvghm	402	10	traction	traction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	11	systems	system	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	402	12	by	by	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	13	gibbs	gibb	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	402	14	sampling	sampling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	402	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	403	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	403	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	403	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	403	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	403	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	403	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	403	7	363–370	363–370	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	403	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	404	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	404	2	Flesch	Flesch	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	404	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	405	1	1948	1948	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	405	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	406	1	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	406	2	new	new	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	406	3	readability	readability	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	406	4	yardstick	yardstick	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	406	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	407	1	Journal	Journal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	407	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	407	3	Applied	Applied	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	407	4	Psychology	Psychology	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	407	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	407	6	32:221	32:221	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	407	7	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	407	8	233	233	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	407	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	408	1	A.C.	A.C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	408	2	Graesser	Graesser	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	408	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	408	4	D.S.	D.S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	409	1	McNamara	McNamara	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	409	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	409	3	M.M.	M.M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	410	1	Louwerse	louwerse	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	410	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	410	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	410	4	Z.	Z.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	410	5	Cai	Cai	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	410	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	411	1	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	411	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	412	1	Coh	Coh	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	412	2	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	412	3	Metrix	Metrix	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	412	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	412	5	Analysis	analysis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	6	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	7	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	9	co-	co-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	412	10	hesion	hesion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	412	12	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	412	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	413	1	Behavior	Behavior	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	413	2	Research	Research	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	413	3	Methods	Methods	NNPS
work_hea3gijzhnc2rl4sxgyj6pvghm	413	4	In-	In-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	413	5	struments	strument	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	413	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	413	7	Computers	computer	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	413	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	413	9	36(2):193–202	36(2):193–202	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	413	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	414	1	H.	H.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	414	2	Ji	Ji	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	414	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	414	4	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	414	5	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	414	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	415	1	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	415	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	416	1	Gender	gender	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	2	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	416	3	animacy	animacy	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	4	knowledge	knowledge	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	5	discovery	discovery	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	6	from	from	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	7	web	web	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	8	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	416	9	scale	scale	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	10	n	n	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	11	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	416	12	grams	gram	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	416	13	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	14	unsupervised	unsupervised	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	416	15	person	person	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	16	name	name	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	17	detection	detection	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	416	18	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	417	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	417	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	417	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	417	4	PACLIC	PACLIC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	417	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	418	1	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	418	2	Klein	Klein	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	418	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	418	4	C.D.	C.D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	419	1	Manning	manning	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	419	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	420	1	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	420	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	421	1	Accurate	accurate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	421	2	unlexical-	unlexical-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	421	3	ized	ized	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	421	4	parsing	parsing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	421	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	422	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	422	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	422	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	422	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	422	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	422	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	422	7	423–430	423–430	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	422	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	423	1	S.M.	S.M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	424	1	Kosslyn	Kosslyn	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	424	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	425	1	1980	1980	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	425	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	426	1	Image	image	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	426	2	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	426	3	mind	mind	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	426	4	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	427	1	Harvard	Harvard	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	427	2	Univer-	Univer-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	427	3	sity	sity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	427	4	Press	Press	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	427	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	428	1	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	428	2	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	428	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	429	1	2003	2003	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	429	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	430	1	Probabilistic	probabilistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	430	2	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	430	3	structuring	structuring	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	430	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	430	5	Experi-	Experi-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	430	6	ments	ment	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	430	7	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	430	8	sentence	sentence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	430	9	ordering	order	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	430	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	431	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	431	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	431	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	431	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	431	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	431	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	431	7	545–552	545–552	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	431	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	432	1	C.	C.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	432	2	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	432	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	432	4	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	432	5	Hovy	Hovy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	432	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	433	1	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	433	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	434	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	434	2	automated	automated	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	434	3	acquisition	acquisition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	5	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	6	signatures	signature	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	434	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	8	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	9	summarization	summarization	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	434	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	435	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	435	2	Proceed-	Proceed-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	435	3	ings	ing	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	435	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	435	5	COLING	coling	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	435	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	435	7	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	435	8	495–501	495–501	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	435	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	436	1	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	2	Lin	Lin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	4	K.	K.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	5	W.	W.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	6	Church	Church	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	8	H.	H.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	9	Ji	Ji	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	10	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	11	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	12	Sekine	Sekine	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	13	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	14	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	15	Yarowsky	Yarowsky	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	16	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	17	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	18	Bergsma	Bergsma	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	19	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	20	K.	K.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	21	Patil	Patil	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	22	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	23	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	24	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	25	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	26	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	27	Lathbury	Lathbury	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	28	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	29	V.	V.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	30	Rao	Rao	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	31	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	32	K.	K.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	33	Dalwani	Dalwani	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	34	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	436	35	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	436	36	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	37	Narsale	Narsale	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	436	38	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	437	1	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	437	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	438	1	New	new	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	438	2	tools	tool	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	438	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	438	4	web-	web-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	438	5	scale	scale	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	438	6	n	n	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	438	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	438	8	grams	gram	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	438	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	439	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	439	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	439	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	439	4	LREC	LREC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	439	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	440	1	N.	N.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	440	2	McIntyre	McIntyre	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	440	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	440	4	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	440	5	Lapata	Lapata	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	440	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	441	1	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	441	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	442	1	Learning	learn	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	442	2	to	to	TO
work_hea3gijzhnc2rl4sxgyj6pvghm	442	3	tell	tell	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	442	4	tales	tale	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	442	5	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	442	6	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	442	7	data	data	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	8	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	442	9	driven	drive	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	10	approach	approach	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	11	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	12	story	story	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	13	generation	generation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	442	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	443	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	443	2	Pro-	Pro-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	443	3	ceedings	ceeding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	443	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	443	5	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	443	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	443	7	IJCNLP	IJCNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	443	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	443	9	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	443	10	217–225	217–225	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	443	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	444	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	444	2	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	444	3	Mooney	Mooney	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	444	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	444	5	L.	L.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	444	6	Roy	Roy	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	444	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	445	1	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	445	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	446	1	Content	content	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	2	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	446	3	based	base	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	4	book	book	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	5	recommending	recommend	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	446	6	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	446	7	learning	learn	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	446	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	9	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	10	categorization	categorization	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	446	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	447	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	447	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	447	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	447	4	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	447	5	fifth	fifth	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	447	6	ACM	ACM	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	447	7	conference	conference	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	447	8	on	on	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	447	9	Digital	Digital	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	447	10	libraries	library	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	447	11	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	447	12	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	447	13	195–204	195–204	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	447	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	448	1	A.	A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	448	2	Nakhimovsky	Nakhimovsky	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	448	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	449	1	1988	1988	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	449	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	450	1	Aspect	aspect	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	450	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	450	3	aspectual	aspectual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	450	4	class	class	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	450	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	450	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	450	7	the	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	450	8	temporal	temporal	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	450	9	structure	structure	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	450	10	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	450	11	narrative	narrative	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	450	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	451	1	Computational	computational	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	451	2	Lin-	Lin-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	451	3	guistics	guistic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	451	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	451	5	14(2):29–43	14(2):29–43	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	451	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	451	7	June	June	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	451	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	452	1	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	2	Pazzani	Pazzani	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	452	4	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	5	Muramatsu	Muramatsu	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	452	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	452	8	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	9	Billsus	Billsus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	452	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	453	1	1996	1996	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	453	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	454	1	Syskill	Syskill	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	454	2	&	&	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	454	3	Webert	webert	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	454	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	454	5	Identifying	identify	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	454	6	interesting	interesting	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	454	7	web	web	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	454	8	sites	site	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	454	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	455	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	455	2	Pro-	Pro-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	455	3	ceedings	ceeding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	455	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	455	5	AAAI	aaai	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	455	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	455	7	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	455	8	54–61	54–61	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	455	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	456	1	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	456	2	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	456	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	456	4	A.	A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	456	5	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	456	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	457	1	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	457	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	458	1	Revisiting	revisit	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	458	2	readabil-	readabil-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	458	3	ity	ity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	458	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	458	5	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	458	6	unified	unified	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	458	7	framework	framework	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	458	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	458	9	predicting	predict	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	458	10	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	458	11	quality	quality	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	458	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	459	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	459	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	459	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	459	4	EMNLP	EMNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	459	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	459	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	459	7	186–195	186–195	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	459	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	460	1	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	460	2	Pitler	Pitler	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	460	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	460	4	A.	A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	460	5	Nenkova	Nenkova	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	460	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	461	1	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	461	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	462	1	Using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	462	2	syntax	syntax	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	462	3	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	462	4	dis-	dis-	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	462	5	ambiguate	ambiguate	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	462	6	explicit	explicit	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	462	7	discourse	discourse	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	462	8	connectives	connective	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	462	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	462	10	text	text	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	462	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	463	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	463	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	463	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	463	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	463	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	463	6	IJCNLP	IJCNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	463	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	463	8	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	463	9	13–16	13–16	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	463	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	464	1	R	R	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	464	2	Development	Development	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	464	3	Core	Core	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	464	4	Team	Team	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	464	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	464	6	2011	2011	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	464	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	465	1	R	r	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	465	2	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	465	3	A	A	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	465	4	Language	Language	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	465	5	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	465	6	Environment	Environment	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	465	7	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	465	8	Statistical	Statistical	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	465	9	Computing	Computing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	465	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	466	1	R	R	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	466	2	Foundation	Foundation	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	466	3	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	466	4	Statistical	Statistical	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	466	5	Computing	Computing	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	466	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	467	1	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	2	Ramage	Ramage	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	467	4	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	5	Hall	Hall	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	467	7	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	8	Nallapati	Nallapati	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	467	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	467	10	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	467	11	C.D.	C.D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	468	1	Manning	manning	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	468	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	469	1	2009	2009	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	469	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	470	1	Labeled	Labeled	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	470	2	lda	lda	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	470	3	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	470	4	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	470	5	supervised	supervise	VBN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	6	topic	topic	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	7	model	model	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	8	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	9	credit	credit	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	10	attribution	attribution	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	11	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	12	multi	multi	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	470	13	-	-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	470	14	labeled	labeled	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	470	15	corpora	corpora	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	470	16	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	471	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	471	2	Proceed-	Proceed-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	471	3	ings	ing	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	471	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	471	5	EMNLP	EMNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	471	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	471	7	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	471	8	248–256	248–256	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	471	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	472	1	A.	A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	472	2	Rozovskaya	Rozovskaya	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	472	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	472	4	D.	D.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	472	5	Roth	Roth	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	472	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	473	1	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	473	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	474	1	Generating	generate	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	474	2	confu-	confu-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	3	sion	sion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	4	sets	set	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	474	5	for	for	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	6	context	context	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	474	8	sensitive	sensitive	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	474	9	error	error	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	10	correction	correction	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	474	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	475	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	475	2	Pro-	Pro-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	475	3	ceedings	ceeding	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	475	4	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	475	5	EMNLP	emnlp	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	475	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	475	7	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	475	8	961–970	961–970	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	475	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	476	1	E.	E.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	476	2	Sandhaus	Sandhaus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	476	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	477	1	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	477	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	478	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	478	2	new	new	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	478	3	york	york	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	478	4	times	time	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	478	5	annotated	annotate	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	478	6	cor-	cor-	RB
work_hea3gijzhnc2rl4sxgyj6pvghm	478	7	pus	pus	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	478	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	479	1	Corpus	corpus	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	479	2	number	number	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	479	3	LDC2008T19	ldc2008t19	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	479	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	479	5	Linguistic	Linguistic	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	479	6	Data	Data	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	479	7	Consortium	Consortium	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	479	8	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	479	9	Philadelphia	Philadelphia	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	479	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	480	1	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	480	2	Schwarm	Schwarm	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	480	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	480	4	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	480	5	Ostendorf	Ostendorf	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	480	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	481	1	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	481	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	482	1	Reading	read	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	482	2	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	3	as-	as-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	4	sessment	sessment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	5	using	use	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	482	6	support	support	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	7	vector	vector	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	8	machines	machine	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	482	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	482	10	statistical	statistical	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	482	11	language	language	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	482	12	models	model	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	482	13	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	483	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	483	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	483	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	483	4	ACL	ACL	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	483	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	483	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	483	7	523	523	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	483	8	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	483	9	530	530	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	483	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	484	1	351	351	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	484	2	H.A.	H.A.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	485	1	Semetko	Semetko	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	485	2	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	485	3	P.M.	P.M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	485	4	Valkenburg	Valkenburg	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	485	5	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	486	1	2000	2000	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	486	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	487	1	Framing	frame	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	487	2	eu-	eu-	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	487	3	ropean	ropean	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	487	4	politics	politic	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	487	5	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	487	6	A	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	487	7	content	content	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	487	8	analysis	analysis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	487	9	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	487	10	press	press	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	487	11	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	487	12	televi-	televi-	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	487	13	sion	sion	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	487	14	news	news	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	487	15	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	488	1	Journal	Journal	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	488	2	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	488	3	communication	communication	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	488	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	488	5	50(2):93–109	50(2):93–109	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	488	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	489	1	V.	V.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	489	2	Spandel	Spandel	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	489	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	490	1	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	490	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	491	1	Creating	create	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	491	2	Writers	writer	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	491	3	Through	through	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	491	4	6-Trait	6-trait	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	491	5	Writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	491	6	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	491	7	Assessment	Assessment	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	491	8	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	491	9	Instruction	Instruction	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	491	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	492	1	Allyn	Allyn	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	2	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	492	3	Ba-	Ba-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	4	con	con	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	492	6	Inc.	Inc.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	7	S.	S.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	8	H.	H.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	9	Stocking	Stocking	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	492	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	493	1	2010	2010	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	493	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	494	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	494	2	New	New	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	3	York	York	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	4	Times	Times	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	5	Reader	Reader	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	6	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	494	7	Sci-	Sci-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	8	ence	ence	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	494	9	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	494	10	Technology	Technology	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	494	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	495	1	CQ	CQ	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	495	2	Press	Press	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	495	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	495	4	Washington	Washington	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	495	5	DC	DC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	495	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	496	1	P.	P.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	496	2	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	1	Stone	Stone	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	2	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	497	3	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	4	Kirsh	Kirsh	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	497	6	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	497	7	Cambridge	Cambridge	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	8	Computer	Computer	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	9	Asso-	Asso-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	497	10	ciates	ciate	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	497	11	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	498	1	1966	1966	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	498	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	499	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	499	2	General	General	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	3	Inquirer	Inquirer	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	499	5	A	A	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	6	Computer	Computer	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	7	Ap-	ap-	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	499	8	proach	proach	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	499	9	to	to	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	499	10	Content	Content	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	11	Analysis	Analysis	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	499	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	500	1	MIT	MIT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	500	2	Press	Press	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	500	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	501	1	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	501	2	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	501	3	Tetreault	Tetreault	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	501	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	501	5	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	501	6	Chodorow	Chodorow	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	501	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	502	1	2008	2008	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	502	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	503	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	503	2	ups	up	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	503	3	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	503	4	downs	down	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	503	5	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	6	preposition	preposition	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	7	error	error	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	8	detection	detection	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	9	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	10	esl	esl	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	503	11	writing	writing	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	503	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	504	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	504	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	504	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	504	4	COLING	COLING	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	504	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	504	6	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	504	7	865–872	865–872	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	504	8	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	505	1	L.	L.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	505	2	von	von	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	505	3	Ahn	Ahn	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	505	4	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	505	5	L.	L.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	505	6	Dabbish	Dabbish	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	505	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	506	1	2004	2004	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	506	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	507	1	Labeling	label	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	507	2	images	image	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	507	3	with	with	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	507	4	a	a	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	507	5	computer	computer	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	507	6	game	game	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	507	7	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	508	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	508	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	508	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	508	4	CHI	CHI	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	508	5	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	508	6	pages	page	NNS
work_hea3gijzhnc2rl4sxgyj6pvghm	508	7	319	319	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	508	8	–	–	:
work_hea3gijzhnc2rl4sxgyj6pvghm	508	9	326	326	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	508	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	509	1	R.	R.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	509	2	L.	L.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	509	3	Weide	Weide	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	509	4	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	510	1	1998	1998	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	510	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	511	1	The	the	DT
work_hea3gijzhnc2rl4sxgyj6pvghm	511	2	cmu	cmu	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	511	3	pronunciation	pronunciation	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	511	4	dictio-	dictio-	VBD
work_hea3gijzhnc2rl4sxgyj6pvghm	511	5	nary	nary	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	511	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	511	7	release	release	VB
work_hea3gijzhnc2rl4sxgyj6pvghm	511	8	0.6	0.6	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	511	9	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	511	10	http://www.speech.cs.cmu.edu/cgi-	http://www.speech.cs.cmu.edu/cgi-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	511	11	bin	bin	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	511	12	/	/	SYM
work_hea3gijzhnc2rl4sxgyj6pvghm	511	13	cmudict	cmudict	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	511	14	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	512	1	T.	T.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	2	Wilson	Wilson	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	3	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	512	4	J.	J.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	5	Wiebe	Wiebe	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	6	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	512	7	and	and	CC
work_hea3gijzhnc2rl4sxgyj6pvghm	512	8	P.	P.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	9	Hoffmann	Hoffmann	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	512	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	513	1	2005	2005	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	513	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	514	1	Recogniz-	Recogniz-	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	514	2	ing	e	VBG
work_hea3gijzhnc2rl4sxgyj6pvghm	514	3	contextual	contextual	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	514	4	polarity	polarity	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	5	in	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	6	phrase	phrase	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	7	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	514	8	level	level	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	9	sentiment	sentiment	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	10	anal-	anal-	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	514	11	ysis	ysis	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	514	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	515	1	In	in	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	515	2	Proceedings	Proceedings	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	515	3	of	of	IN
work_hea3gijzhnc2rl4sxgyj6pvghm	515	4	HLT	HLT	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	515	5	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	515	6	EMNLP	EMNLP	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	515	7	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	515	8	pages	page	VBZ
work_hea3gijzhnc2rl4sxgyj6pvghm	515	9	347–354	347–354	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	515	10	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	516	1	M.	M.	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	516	2	Wilson	Wilson	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	516	3	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	517	1	1988	1988	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	517	2	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	518	1	MRC	MRC	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	518	2	psycholinguistic	psycholinguistic	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	518	3	database	database	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	518	4	:	:	:
work_hea3gijzhnc2rl4sxgyj6pvghm	518	5	Machine	machine	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	518	6	-	-	HYPH
work_hea3gijzhnc2rl4sxgyj6pvghm	518	7	usable	usable	JJ
work_hea3gijzhnc2rl4sxgyj6pvghm	518	8	dictionary	dictionary	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	518	9	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	518	10	version	version	NN
work_hea3gijzhnc2rl4sxgyj6pvghm	518	11	2.00	2.00	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	518	12	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	519	1	Behavior	Behavior	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	519	2	Research	Research	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	519	3	Methods	Methods	NNP
work_hea3gijzhnc2rl4sxgyj6pvghm	519	4	,	,	,
work_hea3gijzhnc2rl4sxgyj6pvghm	519	5	20(1):6–10	20(1):6–10	CD
work_hea3gijzhnc2rl4sxgyj6pvghm	519	6	.	.	.
work_hea3gijzhnc2rl4sxgyj6pvghm	520	1	352	352	CD