id	sid	tid	token	lemma	pos
work_j4jyyr4gnrbelgfykpcumrgui4	1	1	Transactions	transaction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	1	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	1	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	1	4	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	1	5	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	1	6	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	1	7	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	1	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	1	9	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	1	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	1	11	2013	2013	CD
work_j4jyyr4gnrbelgfykpcumrgui4	1	12	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	1	13	75–88	75–88	CD
work_j4jyyr4gnrbelgfykpcumrgui4	1	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	2	1	Action	Action	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	2	2	Editor	Editor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	2	3	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	2	4	Sharon	Sharon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	2	5	Goldwater	Goldwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	2	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	3	1	Submitted	submit	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	3	2	11/2012	11/2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	3	3	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	3	4	Revised	revise	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	3	5	1/2013	1/2013	CD
work_j4jyyr4gnrbelgfykpcumrgui4	3	6	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	3	7	Published	publish	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	3	8	3/2013	3/2013	CD
work_j4jyyr4gnrbelgfykpcumrgui4	3	9	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	4	1	c	c	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	4	2	©	©	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	4	3	2013	2013	CD
work_j4jyyr4gnrbelgfykpcumrgui4	4	4	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	4	5	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	4	6	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	4	7	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	4	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	5	1	An	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	5	2	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	3	Model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	4	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	5	Inducing	induce	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	5	6	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	7	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	8	Grammars	Grammars	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	5	9	Yonatan	Yonatan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	10	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	5	12	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	13	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	14	Department	Department	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	16	Computer	Computer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	17	Science	Science	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	18	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	5	19	University	University	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	21	Illinois	Illinois	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	22	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	23	Urbana	Urbana	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	24	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	5	25	Champaign	Champaign	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	26	201	201	CD
work_j4jyyr4gnrbelgfykpcumrgui4	5	27	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	28	Goodwin	Goodwin	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	29	Ave	Ave	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	30	Urbana	Urbana	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	31	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	5	32	IL	IL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	33	61801	61801	CD
work_j4jyyr4gnrbelgfykpcumrgui4	5	34	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	5	35	bisk1,juliahmr}@illinois.edu	bisk1,juliahmr}@illinois.edu	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	36	Abstract	Abstract	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	37	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	5	38	introduce	introduce	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	5	39	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	5	40	novel	novel	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	41	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	5	42	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	5	43	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	44	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	45	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	5	46	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	47	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	48	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	49	Cat-	Cat-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	50	egorial	egorial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	5	51	Grammars	Grammars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	52	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	5	53	POS	POS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	5	54	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	5	55	tagged	tag	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	5	56	text	text	NN
work_j4jyyr4gnrbelgfykpcumrgui4	5	57	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	6	1	It	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	6	2	achieves	achieve	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	6	3	state	state	NN
work_j4jyyr4gnrbelgfykpcumrgui4	6	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	6	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	6	6	art	art	NN
work_j4jyyr4gnrbelgfykpcumrgui4	6	7	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	6	8	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	6	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	6	10	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	6	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	6	12	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	6	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	6	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	6	15	induces	induce	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	6	16	linguisti-	linguisti-	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	6	17	cally	cally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	6	18	plausible	plausible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	6	19	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	6	20	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	7	1	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	7	2	Introduction	introduction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	7	3	What	what	WP
work_j4jyyr4gnrbelgfykpcumrgui4	7	4	grammatical	grammatical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	7	5	representation	representation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	7	6	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	7	7	appropriate	appropriate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	7	8	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	7	9	unsupervised	unsupervised	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	7	10	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	7	11	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	7	12	?	?	.
work_j4jyyr4gnrbelgfykpcumrgui4	8	1	Initial	initial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	8	2	attempts	attempt	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	8	3	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	8	4	context	context	NN
work_j4jyyr4gnrbelgfykpcumrgui4	8	5	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	8	6	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	8	7	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	8	8	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	8	9	CFGs	CFGs	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	8	10	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	8	11	were	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	8	12	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	8	13	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	8	14	successful	successful	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	8	15	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	8	16	Carroll	Carroll	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	8	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	8	18	Charniak	Charniak	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	8	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	8	20	1992	1992	CD
work_j4jyyr4gnrbelgfykpcumrgui4	8	21	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	8	22	Charniak	Charniak	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	8	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	8	24	1993	1993	CD
work_j4jyyr4gnrbelgfykpcumrgui4	8	25	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	8	26	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	9	1	One	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	9	2	reason	reason	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	3	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	9	4	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	9	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	6	CFGs	CFGs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	9	7	require	require	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	9	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	9	9	specification	specification	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	10	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	9	12	finite	finite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	13	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	15	nonterminal	nonterminal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	16	cat-	cat-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	17	egories	egorie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	9	19	rewrite	rewrite	VB
work_j4jyyr4gnrbelgfykpcumrgui4	9	20	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	9	22	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	9	23	unless	unless	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	24	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	9	25	adopts	adopt	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	9	26	lin-	lin-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	27	guistic	guistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	28	principles	principle	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	29	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	30	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	31	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	32	-	-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	33	bar	bar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	34	theory	theory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	9	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	9	36	Jackendoff	Jackendoff	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	9	37	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	9	38	1977	1977	CD
work_j4jyyr4gnrbelgfykpcumrgui4	9	39	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	9	40	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	9	41	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	9	42	nonterminals	nonterminal	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	43	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	9	44	essentially	essentially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	9	45	arbitrary	arbitrary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	46	labels	label	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	47	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	9	48	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	9	49	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	9	50	combined	combine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	9	51	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	9	52	arbitrary	arbitrary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	9	53	ways	way	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	9	54	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	10	1	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	10	2	further	further	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	10	3	CFG	CFG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	4	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	10	5	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	10	6	approaches	approach	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	10	7	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	10	8	been	be	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	10	9	proposed	propose	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	10	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	10	11	Clark	Clark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	10	13	2001	2001	CD
work_j4jyyr4gnrbelgfykpcumrgui4	10	14	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	10	15	Kurihara	Kurihara	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	10	17	Sato	Sato	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	10	19	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	10	20	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	10	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	10	22	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	10	23	re-	re-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	10	24	cent	cent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	10	25	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	10	26	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	10	27	followed	follow	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	10	28	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	29	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	10	30	Manning	Manning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	31	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	10	32	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	10	33	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	10	34	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	10	35	developing	develop	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	10	36	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	10	37	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	10	38	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	10	39	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	10	40	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	10	41	projec-	projec-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	42	tive	tive	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	10	43	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	10	44	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	10	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	11	1	It	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	11	2	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	11	3	been	be	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	11	4	shown	show	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	11	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	11	6	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	11	7	sophisticated	sophisticated	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	11	8	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	9	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	11	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	11	Headden	Headden	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	12	III	III	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	13	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	11	14	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	15	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	17	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	18	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	11	19	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	20	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	21	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	22	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	24	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	25	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	11	26	Cohen	Cohen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	27	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	11	28	Smith	Smith	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	30	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	31	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	32	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	11	33	learning	learning	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	34	regimes	regime	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	11	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	36	Spitkovsky	Spitkovsky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	37	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	11	38	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	39	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	40	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	41	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	42	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	43	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	44	as	as	RB
work_j4jyyr4gnrbelgfykpcumrgui4	11	45	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	11	46	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	11	47	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	11	48	incorporation	incorporation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	49	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	11	50	prior	prior	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	11	51	lin-	lin-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	11	52	guistic	guistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	11	53	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	54	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	55	Cohen	Cohen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	56	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	11	57	Smith	Smith	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	58	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	59	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	60	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	11	61	Berg-	Berg-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	62	Kirkpatrick	Kirkpatrick	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	63	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	11	64	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	65	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	66	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	67	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	11	68	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	69	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	70	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	71	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	72	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	11	73	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	11	74	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	11	75	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	11	76	lead	lead	VB
work_j4jyyr4gnrbelgfykpcumrgui4	11	77	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	11	78	significant	significant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	11	79	improvement	improvement	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	80	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	11	81	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	82	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	11	83	Manning	Manning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	11	84	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	11	85	baseline	baseline	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	86	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	11	87	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	12	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	12	2	use	use	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	4	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	5	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	12	6	circumvents	circumvent	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	12	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	12	8	question	question	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	10	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	12	11	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	12	12	obtain	obtain	VB
work_j4jyyr4gnrbelgfykpcumrgui4	12	13	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	12	14	appropriate	appropriate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	12	15	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	16	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	17	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	12	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	12	19	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	20	depen-	depen-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	21	dency	dency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	22	parses	parse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	12	23	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	12	24	simply	simply	RB
work_j4jyyr4gnrbelgfykpcumrgui4	12	25	defined	define	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	12	26	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	27	unlabeled	unlabeled	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	12	28	edges	edge	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	12	29	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	30	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	12	31	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	12	32	items	item	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	12	33	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	12	34	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	12	35	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	12	36	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	13	1	But	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	13	2	de-	de-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	3	pendency	pendency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	4	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	5	make	make	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	13	6	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	13	7	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	13	8	difficult	difficult	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	9	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	10	cap-	cap-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	13	11	ture	ture	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	12	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	13	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	14	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	15	structures	structure	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	13	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	13	18	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	13	19	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	13	20	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	13	21	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	13	22	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	13	23	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	13	24	show	show	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	13	25	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	26	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	13	27	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	13	28	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	13	29	advantageous	advantageous	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	30	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	31	refor-	refor-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	32	mulate	mulate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	13	33	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	13	34	underlying	underlie	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	13	35	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	36	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	37	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	38	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	39	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	40	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	13	41	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	42	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	13	43	substitution	substitution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	44	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	45	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	13	46	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	13	47	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	13	48	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	13	49	pairs	pair	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	13	50	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	51	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	52	treelets	treelet	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	53	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	13	54	specify	specify	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	13	55	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	13	56	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	13	57	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	13	58	left	left	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	59	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	13	60	right	right	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	13	61	dependents	dependent	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	13	62	they	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	13	63	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	13	64	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	14	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	14	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	14	3	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	14	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	14	6	explore	explore	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	14	7	yet	yet	RB
work_j4jyyr4gnrbelgfykpcumrgui4	14	8	another	another	DT
work_j4jyyr4gnrbelgfykpcumrgui4	14	9	option	option	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	10	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	14	11	instead	instead	RB
work_j4jyyr4gnrbelgfykpcumrgui4	14	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	14	13	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	14	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	14	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	14	16	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	14	17	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	14	18	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	14	19	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	14	20	Gram-	Gram-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	14	21	mar	mar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	14	23	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	14	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	14	25	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	14	26	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	14	27	1996	1996	CD
work_j4jyyr4gnrbelgfykpcumrgui4	14	28	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	14	29	2000	2000	CD
work_j4jyyr4gnrbelgfykpcumrgui4	14	30	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	14	31	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	14	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	14	33	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	14	34	linguistically	linguistically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	14	35	expressive	expressive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	14	36	formalism	formalism	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	37	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	14	38	pairs	pair	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	14	39	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	14	40	items	item	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	14	41	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	14	42	rich	rich	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	14	43	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	14	44	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	14	45	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	14	46	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	14	47	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	48	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	14	49	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	14	50	in-	in-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	14	51	formation	formation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	14	52	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	15	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	15	2	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	15	3	seem	seem	VB
work_j4jyyr4gnrbelgfykpcumrgui4	15	4	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	15	5	puzzling	puzzling	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	15	6	choice	choice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	15	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	15	8	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	15	9	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	15	10	requires	require	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	15	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	15	12	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	15	13	larger	large	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	15	14	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	15	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	15	16	cat-	cat-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	15	17	egories	egorie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	15	18	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	15	19	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	15	20	commonly	commonly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	15	21	assumed	assume	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	15	22	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	15	23	CFGs	cfg	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	15	24	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	16	1	How-	How-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	16	2	ever	ever	RB
work_j4jyyr4gnrbelgfykpcumrgui4	16	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	16	4	unlike	unlike	IN
work_j4jyyr4gnrbelgfykpcumrgui4	16	5	CFG	CFG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	16	6	nonterminals	nonterminal	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	16	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	16	8	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	16	9	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	16	10	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	16	11	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	16	12	arbitrary	arbitrary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	16	13	symbols	symbol	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	16	14	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	16	15	they	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	16	16	encode	encode	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	16	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	16	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	16	19	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	16	20	deter-	deter-	CD
work_j4jyyr4gnrbelgfykpcumrgui4	16	21	mined	mine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	16	22	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	16	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	16	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	16	25	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	16	26	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	16	27	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	16	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	16	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	16	30	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	16	31	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	16	32	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	16	33	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	16	34	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	16	35	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	16	36	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	16	37	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	16	38	takes	take	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	16	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	17	1	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	17	2	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	17	3	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	17	4	similar	similar	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	17	5	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	17	6	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	17	7	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	17	8	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	17	9	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	17	10	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	17	11	pairs	pair	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	17	12	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	17	13	items	item	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	17	14	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	17	15	rich	rich	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	17	16	items	item	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	17	17	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	17	18	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	17	19	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	17	20	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	17	21	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	17	22	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	17	23	in-	in-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	17	24	formation	formation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	17	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	18	1	Like	like	IN
work_j4jyyr4gnrbelgfykpcumrgui4	18	2	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	18	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	18	4	projective	projective	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	18	5	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	18	6	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	18	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	18	8	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	18	9	restrict	restrict	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	18	10	ourselves	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	18	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	18	12	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	18	13	weakly	weakly	NN
work_j4jyyr4gnrbelgfykpcumrgui4	18	14	context-	context-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	18	15	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	18	16	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	18	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	18	18	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	18	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	19	1	But	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	19	2	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	19	3	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	19	4	does	do	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	19	5	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	19	6	dis-	dis-	VB
work_j4jyyr4gnrbelgfykpcumrgui4	19	7	tinguish	tinguish	NN
work_j4jyyr4gnrbelgfykpcumrgui4	19	8	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	19	9	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	19	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	19	11	modifier	modifi	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	19	12	dependen-	dependen-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	19	13	cies	cie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	19	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	19	15	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	19	16	makes	make	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	19	17	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	19	18	explicit	explicit	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	19	19	distinction	distinction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	19	20	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	19	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	19	22	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	19	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	20	1	And	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	20	2	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	20	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	20	4	elementary	elementary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	20	5	trees	tree	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	20	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	20	7	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	20	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	20	9	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	20	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	20	11	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	20	12	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	20	13	’s	’	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	20	14	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	20	15	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	20	16	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	20	17	internal	internal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	20	18	nodel	nodel	NN
work_j4jyyr4gnrbelgfykpcumrgui4	20	19	la-	la-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	20	20	bels	bel	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	20	21	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	20	22	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	20	23	obvious	obvious	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	20	24	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	20	25	interpretation	interpretation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	20	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	20	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	20	28	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	20	29	behavior	behavior	NN
work_j4jyyr4gnrbelgfykpcumrgui4	20	30	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	20	31	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	20	32	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	20	33	constituent	constituent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	20	34	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	20	35	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	20	36	directly	directly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	20	37	inferred	infer	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	20	38	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	20	39	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	20	40	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	20	41	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	21	1	To	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	21	2	see	see	VB
work_j4jyyr4gnrbelgfykpcumrgui4	21	3	whether	whether	IN
work_j4jyyr4gnrbelgfykpcumrgui4	21	4	75	75	CD
work_j4jyyr4gnrbelgfykpcumrgui4	21	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	21	6	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	21	7	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	21	8	identified	identify	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	21	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	21	10	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	21	11	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	21	12	prop-	prop-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	21	13	erties	ertie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	21	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	21	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	21	16	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	21	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	21	18	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	21	19	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	21	20	hence	hence	RB
work_j4jyyr4gnrbelgfykpcumrgui4	21	21	sufficient	sufficient	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	21	22	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	21	23	in-	in-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	21	24	spect	spect	VB
work_j4jyyr4gnrbelgfykpcumrgui4	21	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	21	26	induced	induced	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	21	27	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	21	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	22	1	Conversely	conversely	RB
work_j4jyyr4gnrbelgfykpcumrgui4	22	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	22	3	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	22	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	22	5	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	22	6	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	22	7	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	22	8	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	22	9	show	show	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	22	10	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	22	11	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	22	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	22	13	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	22	14	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	22	15	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	22	16	properties	property	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	22	17	makes	make	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	22	18	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	22	19	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	22	20	easy	easy	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	22	21	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	22	22	cre-	cre-	VB
work_j4jyyr4gnrbelgfykpcumrgui4	22	23	ate	ate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	22	24	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	22	25	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	22	26	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	22	27	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	22	28	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	22	29	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	22	30	accurate	accurate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	22	31	unsu-	unsu-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	22	32	pervised	pervise	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	22	33	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	22	34	parsing	parsing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	22	35	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	23	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	23	2	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	23	3	recently	recently	RB
work_j4jyyr4gnrbelgfykpcumrgui4	23	4	proposed	propose	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	23	5	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	23	6	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	23	7	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	23	8	inducing	induce	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	23	9	CCGs	CCGs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	23	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	23	11	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	23	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	23	13	Hocken-	Hocken-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	23	14	maier	maier	NN
work_j4jyyr4gnrbelgfykpcumrgui4	23	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	23	16	2012b	2012b	LS
work_j4jyyr4gnrbelgfykpcumrgui4	23	17	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	23	18	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	23	19	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	23	20	been	be	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	23	21	shown	show	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	23	22	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	23	23	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	23	24	competitive	competitive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	23	25	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	23	26	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	23	27	approaches	approach	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	23	28	even	even	RB
work_j4jyyr4gnrbelgfykpcumrgui4	23	29	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	23	30	paired	pair	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	23	31	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	23	32	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	23	33	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	23	34	simple	simple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	23	35	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	23	36	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	23	37	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	23	38	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	23	39	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	23	40	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	23	41	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	23	42	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	23	43	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	23	44	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	23	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	24	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	24	3	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	24	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	24	6	pair	pair	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	24	7	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	24	8	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	9	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	10	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	24	12	novel	novel	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	13	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	14	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	15	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	16	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	24	17	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	24	18	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	24	19	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	20	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	24	21	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	22	factorization	factorization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	24	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	24	25	derivations	derivation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	24	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	24	27	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	24	28	show	show	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	24	29	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	30	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	24	31	outperforms	outperform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	24	32	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	24	33	original	original	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	34	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	35	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	24	36	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	37	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	38	approaches	approach	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	24	39	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	40	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	24	41	large	large	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	24	42	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	43	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	24	44	lan-	lan-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	24	45	guages	guage	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	24	46	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	25	1	Our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	25	2	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	25	3	indicate	indicate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	25	4	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	25	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	25	6	use	use	NN
work_j4jyyr4gnrbelgfykpcumrgui4	25	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	25	8	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	25	9	yields	yield	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	25	10	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	25	11	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	25	12	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	25	13	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	25	14	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	25	15	robust	robust	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	25	16	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	25	17	dealing	deal	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	25	18	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	25	19	longer	long	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	25	20	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	25	21	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	25	22	most	most	JJS
work_j4jyyr4gnrbelgfykpcumrgui4	25	23	de-	de-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	25	24	pendency	pendency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	25	25	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	25	26	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	25	27	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	25	28	approaches	approach	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	25	29	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	26	1	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	26	2	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	3	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	4	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	5	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	6	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	7	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	8	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	26	9	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	26	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	26	11	2000	2000	CD
work_j4jyyr4gnrbelgfykpcumrgui4	26	12	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	26	13	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	26	14	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	26	15	linguistically	linguistically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	26	16	expressive	expressive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	26	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	26	18	lexicalized	lexicalize	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	26	19	grammar	grammar	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	26	20	formalism	formalism	NN
work_j4jyyr4gnrbelgfykpcumrgui4	26	21	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	26	22	associates	associate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	26	23	rich	rich	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	26	24	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	26	25	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	26	26	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	26	27	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	26	28	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	26	29	constituents	constituent	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	26	30	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	27	1	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	27	2	simplicity	simplicity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	27	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	27	4	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	27	5	restrict	restrict	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	27	6	ourselves	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	27	7	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	27	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	27	9	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	27	10	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	27	11	atomic	atomic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	27	12	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	27	13	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	27	14	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	27	15	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	27	16	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	27	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	27	18	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	27	19	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	27	20	encompassing	encompass	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	27	21	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	27	22	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	27	23	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	27	24	noun	noun	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	27	25	phrases	phrase	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	27	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	27	27	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	27	28	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	27	29	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	27	30	recursively	recursively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	27	31	build	build	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	27	32	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	27	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	28	1	Complex	complex	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	28	2	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	28	3	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	28	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	28	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	28	6	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	7	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	28	8	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	28	9	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	10	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	28	11	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	28	13	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	28	14	represent	represent	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	28	15	functions	function	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	28	16	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	28	17	return	return	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	28	18	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	28	19	result	result	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	28	21	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	22	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	23	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	28	24	combined	combine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	28	25	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	28	26	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	28	27	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	28	29	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	28	30	Y.	Y.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	29	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	29	2	directionality	directionality	NN
work_j4jyyr4gnrbelgfykpcumrgui4	29	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	29	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	29	5	slash	slash	NN
work_j4jyyr4gnrbelgfykpcumrgui4	29	6	indicates	indicate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	29	7	whether	whether	IN
work_j4jyyr4gnrbelgfykpcumrgui4	29	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	29	9	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	29	10	precedes	precede	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	29	11	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	29	12	follows	follow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	29	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	29	14	functor	functor	NN
work_j4jyyr4gnrbelgfykpcumrgui4	29	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	30	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	30	2	write	write	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	30	3	X|Y	X|Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	30	4	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	30	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	30	6	direction	direction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	30	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	30	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	30	9	slash	slash	NN
work_j4jyyr4gnrbelgfykpcumrgui4	30	10	does	do	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	30	11	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	30	12	matter	matter	VB
work_j4jyyr4gnrbelgfykpcumrgui4	30	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	31	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	31	2	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	31	3	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	31	4	encodes	encode	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	31	5	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	31	6	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	31	7	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	31	8	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	31	9	information	information	NN
work_j4jyyr4gnrbelgfykpcumrgui4	31	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	32	1	It	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	32	2	pairs	pair	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	32	3	every	every	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	4	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	5	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	6	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	7	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	9	cate-	cate-	FW
work_j4jyyr4gnrbelgfykpcumrgui4	32	10	gories	gorie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	32	11	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	32	12	define	define	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	32	13	both	both	CC
work_j4jyyr4gnrbelgfykpcumrgui4	32	14	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	32	15	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	16	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	17	behavior	behavior	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	18	as	as	RB
work_j4jyyr4gnrbelgfykpcumrgui4	32	19	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	32	20	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	22	overall	overall	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	23	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	24	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	25	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	26	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	27	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	28	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	29	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	30	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	31	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	32	he	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	32	33	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	34	girl	girl	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	35	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	36	lunch	lunch	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	37	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	38	...	...	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	39	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	40	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	41	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	32	42	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	43	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	44	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	45	good	good	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	46	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	47	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	48	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	49	eating	eat	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	32	50	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	51	...	...	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	32	52	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	53	S\N	s\n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	54	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	55	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	56	sleeps	sleep	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	32	57	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	58	ate	eat	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	32	59	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	60	eating	eat	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	32	61	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	62	...	...	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	63	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	64	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	65	S\N)/N	s\n)/n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	66	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	67	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	68	sees	see	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	32	69	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	70	ate	eat	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	32	71	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	72	...	...	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	32	73	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	74	S\S	s\s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	75	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	76	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	77	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	32	78	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	79	today	today	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	80	...	...	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	32	81	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	82	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	83	S\N)/(S\N	S\N)/(S\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	32	84	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	85	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	86	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	87	good	good	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	88	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	89	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	90	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	91	...	...	:
work_j4jyyr4gnrbelgfykpcumrgui4	32	92	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	32	93	To	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	32	94	draw	draw	VB
work_j4jyyr4gnrbelgfykpcumrgui4	32	95	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	96	simple	simple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	97	contrast	contrast	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	98	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	32	99	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	100	Spanish	Spanish	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	32	101	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	32	102	would	would	MD
work_j4jyyr4gnrbelgfykpcumrgui4	32	103	expect	expect	VB
work_j4jyyr4gnrbelgfykpcumrgui4	32	104	adjectives	adjective	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	32	105	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	32	106	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	32	107	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	108	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	109	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	32	110	because	because	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	111	Spanish	spanish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	32	112	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	113	ordering	ordering	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	114	dictates	dictate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	32	115	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	32	116	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	117	adjective	adjective	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	118	fol-	fol-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	119	low	low	PDT
work_j4jyyr4gnrbelgfykpcumrgui4	32	120	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	32	121	noun	noun	NN
work_j4jyyr4gnrbelgfykpcumrgui4	32	122	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	33	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	33	2	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	33	3	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	33	4	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	33	5	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	33	6	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	7	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	33	8	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	9	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	33	10	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	33	11	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	12	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	33	13	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	14	relations	relation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	33	15	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	33	16	captured	capture	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	33	17	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	33	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	33	19	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	33	20	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	21	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	33	22	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	33	23	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	33	24	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	33	25	e.g.	e.g.	RB
work_j4jyyr4gnrbelgfykpcumrgui4	34	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	34	2	S\N)/N	S\N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	34	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	34	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	34	5	whereas	whereas	IN
work_j4jyyr4gnrbelgfykpcumrgui4	34	6	head	head	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	34	7	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	34	8	modifier	modifier	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	9	relations	relation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	34	10	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	34	11	cap-	cap-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	34	12	tured	ture	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	34	13	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	34	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	34	15	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	34	16	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	34	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	34	19	modifier	modifier	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	34	21	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	34	22	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	34	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	34	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	34	25	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	26	X\X	x\x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	27	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	34	28	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	34	29	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	34	30	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	34	31	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	34	32	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	34	33	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	34	34	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	34	35	further	further	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	34	36	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	34	37	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	34	38	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	34	39	own	own	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	34	40	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	35	1	Our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	35	2	goal	goal	NN
work_j4jyyr4gnrbelgfykpcumrgui4	35	3	will	will	MD
work_j4jyyr4gnrbelgfykpcumrgui4	35	4	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	35	5	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	35	6	automati-	automati-	VB
work_j4jyyr4gnrbelgfykpcumrgui4	35	7	cally	cally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	35	8	learn	learn	VB
work_j4jyyr4gnrbelgfykpcumrgui4	35	9	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	35	10	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	35	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	35	12	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	35	13	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	35	14	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	35	15	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	35	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	36	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	36	2	Figure	figure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	36	3	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	36	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	36	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	36	6	juxtapose	juxtapose	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	36	7	several	several	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	36	8	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	36	9	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	36	10	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	36	11	were	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	36	12	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	36	13	discovered	discover	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	36	14	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	36	15	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	36	16	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	36	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	37	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	37	2	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	4	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	37	5	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	37	6	defined	define	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	37	7	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	8	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	37	9	small	small	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	10	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	13	combinatory	combinatory	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	14	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	37	16	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	37	17	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	37	18	traditionally	traditionally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	37	19	writ-	writ-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	37	20	ten	ten	CD
work_j4jyyr4gnrbelgfykpcumrgui4	37	21	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	22	schemas	schema	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	23	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	37	24	define	define	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	37	25	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	37	26	constituents	constituent	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	27	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	37	28	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	37	29	combined	combine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	37	30	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	31	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	37	32	bottom	bottom	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	33	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	37	34	up	up	RP
work_j4jyyr4gnrbelgfykpcumrgui4	37	35	fashion	fashion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	36	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	37	37	although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	38	genera-	genera-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	39	tive	tive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	40	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	41	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	42	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	43	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	37	44	view	view	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	37	45	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	37	46	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	47	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	37	48	top-	top-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	49	down	down	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	50	manner	manner	NN
work_j4jyyr4gnrbelgfykpcumrgui4	37	51	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	37	52	akin	akin	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	37	53	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	37	54	CFG	CFG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	37	55	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	37	56	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	37	57	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	38	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	38	2	first	first	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	38	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	38	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	38	5	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	38	6	obvious	obvious	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	38	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	38	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	38	9	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	38	10	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	38	11	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	38	12	function	function	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	13	application	application	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	14	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	38	15	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	16	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	38	17	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	18	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	19	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	38	20	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	21	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	38	22	B0	b0	RB
work_j4jyyr4gnrbelgfykpcumrgui4	38	23	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	38	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	38	25	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	26	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	27	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	38	28	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	29	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	38	30	B0	B0	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	31	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	38	32	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	38	33	Here	here	RB
work_j4jyyr4gnrbelgfykpcumrgui4	38	34	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	38	35	functor	functor	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	36	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	37	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	38	38	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	39	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	38	40	X\Y	X\Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	41	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	38	42	applied	apply	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	38	43	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	38	44	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	38	45	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	38	46	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	38	47	resulting	result	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	38	48	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	38	49	X.	X.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	1	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	2	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	3	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	4	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	6	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	8	additional	additional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	9	combinatory	combinatory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	10	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	39	11	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	12	type-	type-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	13	raising	raising	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	15	generalized	generalized	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	16	variants	variant	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	39	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	18	composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	19	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	20	substitution	substitution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	21	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	22	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	39	23	increase	increase	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	39	24	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	39	25	generative	generative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	26	capacity	capacity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	27	be-	be-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	28	yond	yond	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	29	context	context	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	30	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	39	31	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	32	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	39	33	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	34	allow	allow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	39	35	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	36	elegant	elegant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	37	treatment	treatment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	38	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	39	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	40	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	41	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	42	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	39	43	arising	arise	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	39	44	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	45	ex-	ex-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	46	traction	traction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	47	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	48	coordination	coordination	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	49	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	50	scrambling	scrambling	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	51	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	52	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	39	53	follow	follow	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	39	54	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	55	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	56	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	57	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	58	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	59	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	60	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	61	use	use	VB
work_j4jyyr4gnrbelgfykpcumrgui4	39	62	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	63	restricted	restrict	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	39	64	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	65	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	66	without	without	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	67	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	68	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	39	69	raising	raising	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	70	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	71	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	39	72	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	73	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	39	74	ba-	ba-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	75	sic	sic	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	76	composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	77	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	78	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	79	context	context	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	80	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	39	81	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	82	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	39	83	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	84	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	85	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	86	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	87	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	88	Z	Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	89	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	90	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	91	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	92	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	93	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	94	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	95	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	96	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	97	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	98	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	99	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	100	Y\Z	y\z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	101	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	102	X\Z	X\Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	103	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	104	B1X	B1X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	105	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	106	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	107	Y\Z	Y\Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	108	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	109	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	110	X\Z	X\Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	111	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	112	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	113	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	114	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	115	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	116	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	117	Z	Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	118	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	119	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	120	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	121	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	39	122	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	123	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	124	B1X	b1x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	125	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	39	126	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	39	127	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	128	superscript	superscript	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	129	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	39	130	denotes	denote	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	131	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	132	arity	arity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	133	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	134	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	135	compo-	compo-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	136	sition	sition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	137	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	39	138	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	139	too	too	RB
work_j4jyyr4gnrbelgfykpcumrgui4	39	140	low	low	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	141	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	39	142	recover	recover	VB
work_j4jyyr4gnrbelgfykpcumrgui4	39	143	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	144	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	145	projective	projective	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	39	146	de-	de-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	147	pendencies	pendencie	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	148	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	39	149	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	39	150	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	39	151	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	152	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	39	153	thus	thus	RB
work_j4jyyr4gnrbelgfykpcumrgui4	39	154	weakly	weakly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	39	155	equiva-	equiva-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	156	lent	lend	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	39	157	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	158	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	39	159	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	160	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	39	161	representations	representation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	39	162	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	39	163	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	39	164	commonly	commonly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	39	165	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	39	166	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	39	167	grammar	grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	168	induction	induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	39	169	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	40	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	40	2	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	40	3	role	role	NN
work_j4jyyr4gnrbelgfykpcumrgui4	40	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	40	5	composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	40	6	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	40	7	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	40	8	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	40	9	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	40	10	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	40	11	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	40	12	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	40	13	sentential	sentential	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	40	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	40	15	verb	verb	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	40	16	modifiers	modifier	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	40	17	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	40	18	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	40	19	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	40	20	cat-	cat-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	40	21	egories	egorie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	40	22	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	40	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	40	24	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	40	25	S\S	S\S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	40	26	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	40	27	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	40	28	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	40	29	S.	S.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	1	Composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	2	in-	in-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	3	76	76	CD
work_j4jyyr4gnrbelgfykpcumrgui4	41	4	troduces	troduce	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	41	5	spurious	spurious	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	6	ambiguities	ambiguity	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	41	8	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	41	9	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	41	10	eliminate	eliminate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	41	11	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	41	12	using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	41	13	Eisner	Eisner	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	14	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	15	1996	1996	CD
work_j4jyyr4gnrbelgfykpcumrgui4	41	16	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	17	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	41	18	normal	normal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	19	form.1	form.1	,
work_j4jyyr4gnrbelgfykpcumrgui4	41	20	Coordinating	coordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	41	21	conjunctions	conjunction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	22	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	41	23	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	41	24	special	special	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	25	cate-	cate-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	26	gory	gory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	27	conj	conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	28	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	41	29	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	41	30	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	41	31	binarize	binarize	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	41	32	coordination	coordination	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	33	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	41	34	follows	follow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	41	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	36	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	37	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	41	38	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	39	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	41	40	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	41	41	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	42	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	41	43	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	44	X[conj	X[conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	45	]	]	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	46	⇒&1	⇒&1	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	47	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	48	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	49	&	&	CC
work_j4jyyr4gnrbelgfykpcumrgui4	41	50	1	1	LS
work_j4jyyr4gnrbelgfykpcumrgui4	41	51	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	52	conj	conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	53	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	54	⇒&2	⇒&2	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	55	X[conj	X[conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	56	]	]	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	57	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	58	&	&	CC
work_j4jyyr4gnrbelgfykpcumrgui4	41	59	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	41	60	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	41	61	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	41	62	Category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	63	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	64	Unlike	unlike	IN
work_j4jyyr4gnrbelgfykpcumrgui4	41	65	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	41	66	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	67	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	41	68	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	41	69	requires	require	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	41	70	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	41	71	in-	in-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	72	ventory	ventory	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	73	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	41	74	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	41	75	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	41	76	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	42	1	Given	give	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	42	2	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	42	3	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	42	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	42	5	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	42	6	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	42	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	42	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	42	9	combinatory	combinatory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	42	10	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	42	11	define	define	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	42	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	42	13	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	42	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	42	15	parses	parse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	42	16	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	42	17	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	42	18	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	42	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	43	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	43	2	follow	follow	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	43	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	43	4	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	43	5	proposed	propose	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	43	6	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	43	7	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	43	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	43	9	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	43	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	43	11	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	43	12	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	43	13	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	43	14	au-	au-	VB
work_j4jyyr4gnrbelgfykpcumrgui4	43	15	tomatically	tomatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	43	16	induce	induce	VB
work_j4jyyr4gnrbelgfykpcumrgui4	43	17	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	43	18	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	43	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	44	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	2	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	44	4	initialized	initialize	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	44	5	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	6	pairing	pair	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	44	7	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	8	nominal	nominal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	44	9	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	44	11	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	13	pro-	pro-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	14	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	15	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	44	16	determiners	determiner	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	17	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	44	18	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	20	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	21	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	44	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	23	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	24	verb	verb	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	44	25	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	26	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	28	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	29	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	31	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	44	32	coordinating	coordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	44	33	conjunc-	conjunc-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	44	34	tions	tion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	35	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	36	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	37	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	38	conj	conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	39	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	44	40	CONJ	CONJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	41	→	→	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	44	42	conj	conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	43	DET	DET	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	44	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	45	NOUN	NOUN	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	46	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	47	NUM	NUM	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	48	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	49	PRON	PRON	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	50	→	→	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	44	51	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	52	VERB	VERB	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	53	→	→	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	44	54	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	55	Although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	56	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	44	57	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	58	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	44	59	defined	define	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	44	60	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	61	corpus-	corpus-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	62	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	44	63	POS	POS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	64	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	44	65	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	44	66	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	44	67	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	44	68	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	44	69	slightly	slightly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	44	70	modified	modify	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	44	71	version	version	NN
work_j4jyyr4gnrbelgfykpcumrgui4	44	72	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	44	73	Petrov	Petrov	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	74	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	75	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	44	76	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	45	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	45	2	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	45	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	45	4	’s	’s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	45	5	Universal	Universal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	45	6	POS	POS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	45	7	tagset	tagset	NN
work_j4jyyr4gnrbelgfykpcumrgui4	45	8	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	45	9	cat-	cat-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	45	10	egorize	egorize	VB
work_j4jyyr4gnrbelgfykpcumrgui4	45	11	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	45	12	into	into	IN
work_j4jyyr4gnrbelgfykpcumrgui4	45	13	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	45	14	broad	broad	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	45	15	classes	class	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	45	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	46	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	46	2	primary	primary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	46	3	changes	change	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	46	4	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	46	5	make	make	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	46	6	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	46	7	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	46	8	mappings	mapping	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	46	9	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	46	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	46	11	addition	addition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	46	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	46	13	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	46	14	distinction	distinction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	46	15	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	46	16	where	where	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	46	17	possible	possible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	46	18	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	46	19	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	46	20	subordi-	subordi-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	46	21	nating	nating	NN
work_j4jyyr4gnrbelgfykpcumrgui4	46	22	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	46	23	coordinating	coordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	46	24	conjunctions	conjunction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	46	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	46	26	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	46	27	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	46	28	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	46	29	auxiliary	auxiliary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	46	30	verbs2	verbs2	NN
work_j4jyyr4gnrbelgfykpcumrgui4	46	31	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	47	1	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	47	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	47	3	initial	initial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	4	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	47	5	consists	consist	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	47	6	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	47	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	47	8	atomic	atomic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	9	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	47	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	47	11	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	47	12	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	47	13	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	47	14	parse	parse	VB
work_j4jyyr4gnrbelgfykpcumrgui4	47	15	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	47	16	complex	complex	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	17	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	47	18	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	47	19	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	47	20	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	47	21	ate	eat	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	47	22	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	47	23	DT	DT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	24	NNS	NNS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	25	VBD	VBD	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	26	RB	RB	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	27	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	47	28	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	29	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	47	30	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	47	31	Complex	complex	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	32	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	33	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	47	34	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	47	35	induced	induce	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	47	36	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	47	37	con-	con-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	47	38	sidering	sidere	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	47	39	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	47	40	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	47	41	context	context	NN
work_j4jyyr4gnrbelgfykpcumrgui4	47	42	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	47	43	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	47	44	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	47	45	appear	appear	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	47	46	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	48	1	Given	give	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	48	2	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	3	input	input	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	4	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	48	6	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	48	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	8	current	current	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	9	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	10	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	48	11	assigns	assign	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	48	12	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	13	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	14	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	15	least	least	JJS
work_j4jyyr4gnrbelgfykpcumrgui4	48	16	some	some	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	19	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	20	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	22	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	48	24	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	48	25	apply	apply	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	48	26	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	27	following	follow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	48	28	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	48	29	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	30	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	48	31	add	add	VB
work_j4jyyr4gnrbelgfykpcumrgui4	48	32	new	new	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	33	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	34	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	35	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	36	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	37	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	48	38	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	39	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	40	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	41	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	48	42	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	43	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	44	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	45	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	48	46	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	48	47	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	48	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	49	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	50	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	51	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	48	52	conj	conj	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	48	53	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	48	54	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	48	55	immediately	immediately	RB
work_j4jyyr4gnrbelgfykpcumrgui4	48	56	adjacent	adjacent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	57	1The	1the	CD
work_j4jyyr4gnrbelgfykpcumrgui4	48	58	normal	normal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	48	59	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	48	60	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	61	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	62	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	48	63	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	48	64	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	48	65	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	48	66	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	48	67	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	48	68	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	48	69	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	48	70	required	require	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	48	71	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	72	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	48	73	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	48	74	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	48	75	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	48	76	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	49	1	2This	2this	CD
work_j4jyyr4gnrbelgfykpcumrgui4	49	2	distinction	distinction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	49	3	was	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	49	4	suggested	suggest	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	49	5	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	49	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	49	7	authors	author	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	49	8	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	49	9	p.c	p.c	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	49	10	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	49	11	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	50	1	Ns	ns	RB
work_j4jyyr4gnrbelgfykpcumrgui4	50	2	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	50	3	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	50	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	51	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	2	modifier	modifi	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	51	3	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	4	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	51	5	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	6	to-	to-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	7	ken	ken	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	8	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	51	9	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	51	10	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	51	11	coordinating	coordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	51	12	conjunctions	conjunction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	51	13	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	51	14	ap-	ap-	CD
work_j4jyyr4gnrbelgfykpcumrgui4	51	15	pear	pear	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	16	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	51	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	18	middle	middle	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	51	20	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	51	21	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	51	22	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	51	23	modify	modify	VB
work_j4jyyr4gnrbelgfykpcumrgui4	51	24	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	25	imme-	imme-	FW
work_j4jyyr4gnrbelgfykpcumrgui4	51	26	diate	diate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	51	27	neighbor	neighbor	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	28	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	51	29	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	51	30	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	31	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	32	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	33	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	51	34	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	35	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	51	36	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	51	37	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	51	38	modifier	modifier	NN
work_j4jyyr4gnrbelgfykpcumrgui4	51	39	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	51	40	S|S	S|S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	41	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	51	42	N|N	N|N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	51	43	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	51	44	itself	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	51	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	52	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	52	2	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	52	3	ate	eat	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	52	4	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	52	5	DT	DT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	6	NNS	NNS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	7	VBD	VBD	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	8	RB	RB	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	9	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	52	10	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	52	11	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	12	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	52	14	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	15	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	52	16	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	17	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	52	19	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	52	20	S\S	s\s	CD
work_j4jyyr4gnrbelgfykpcumrgui4	52	21	S\N	s\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	52	22	These	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	52	23	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	52	24	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	52	25	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	52	26	applied	apply	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	52	27	iteratively	iteratively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	52	28	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	52	29	form	form	VB
work_j4jyyr4gnrbelgfykpcumrgui4	52	30	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	52	31	complex	complex	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	52	32	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	52	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	53	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	53	2	restrict	restrict	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	53	3	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	53	4	cate-	cate-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	53	5	gories	gorie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	53	6	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	53	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	53	8	maximal	maximal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	53	9	arity	arity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	53	10	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	53	11	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	53	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	53	13	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	53	14	disallow	disallow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	53	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	53	16	cat-	cat-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	53	17	egory	egory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	53	18	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	53	19	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	53	20	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	53	21	N)\N	n)\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	53	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	53	23	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	53	24	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	53	25	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	53	26	equivalent	equivalent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	53	27	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	53	28	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	53	29	S\N)/N.	S\N)/N.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	54	2	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	54	3	ate	eat	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	54	4	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	54	5	DT	DT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	6	NNS	NNS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	7	VBD	VBD	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	8	RB	RB	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	9	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	54	10	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	11	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	13	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	15	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	16	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	17	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	18	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	20	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	22	S\S	s\s	CD
work_j4jyyr4gnrbelgfykpcumrgui4	54	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	24	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	25	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	26	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	27	S)/(S	S)/(S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	28	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	29	S)(N\N)/(N\N	S)(N\N)/(N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	30	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	31	S\N	S\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	32	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	33	N\N)\(N\N	N\N)\(N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	34	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	36	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	37	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	38	N)\(N	N)\(N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	39	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	40	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	41	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	42	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	43	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	44	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	45	S)\(S	S)\(S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	46	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	54	47	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	54	48	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	49	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	50	S\S)/(S\S	S\S)/(S\S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	51	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	54	52	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	54	53	resultant	resultant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	54	54	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	55	overly	overly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	54	56	general	general	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	54	57	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	54	58	lexicon	lexicon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	54	59	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	54	60	then	then	RB
work_j4jyyr4gnrbelgfykpcumrgui4	54	61	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	54	62	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	54	63	parse	parse	VB
work_j4jyyr4gnrbelgfykpcumrgui4	54	64	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	54	65	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	54	66	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	54	67	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	55	1	Each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	2	complete	complete	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	55	3	parse	parse	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	4	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	55	5	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	55	6	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	55	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	8	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	9	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	55	10	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	55	11	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	55	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	55	13	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	15	constraint	constraint	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	16	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	55	17	sentences	sentence	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	55	18	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	55	19	contain	contain	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	55	20	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	21	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	55	22	verb	verb	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	23	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	55	24	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	55	25	form	form	VB
work_j4jyyr4gnrbelgfykpcumrgui4	55	26	parses	parse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	55	27	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	28	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	29	S.	S.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	55	30	4	4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	55	31	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	32	new	new	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	55	33	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	34	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	35	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	36	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	55	37	Generative	Generative	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	55	38	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	55	39	define	define	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	55	40	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	41	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	42	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	43	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	44	parse	parse	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	45	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	46	τ	τ	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	47	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	48	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	55	49	product	product	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	50	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	55	51	individual	individual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	55	52	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	53	probabili-	probabili-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	55	54	ties	tie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	55	55	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	56	1	Our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	56	2	previous	previous	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	56	3	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	4	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	5	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	56	6	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	56	7	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	56	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	9	2012b	2012b	LS
work_j4jyyr4gnrbelgfykpcumrgui4	56	10	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	11	uses	use	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	56	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	56	13	most	most	JJS
work_j4jyyr4gnrbelgfykpcumrgui4	56	14	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	56	15	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	16	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	56	17	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	56	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	56	19	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	56	20	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	21	2002	2002	CD
work_j4jyyr4gnrbelgfykpcumrgui4	56	22	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	24	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	56	25	first	first	RB
work_j4jyyr4gnrbelgfykpcumrgui4	56	26	generates	generate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	56	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	56	28	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	29	direction	direction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	30	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	31	left	leave	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	56	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	33	right	right	UH
work_j4jyyr4gnrbelgfykpcumrgui4	56	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	35	unary	unary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	56	36	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	37	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	56	38	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	56	39	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	56	40	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	41	followed	follow	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	56	42	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	56	43	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	56	44	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	45	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	46	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	56	47	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	56	48	finally	finally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	56	49	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	56	50	sister	sister	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	51	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	56	52	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	57	1	3	3	LS
work_j4jyyr4gnrbelgfykpcumrgui4	57	2	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	57	3	factorization	factorization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	57	4	does	do	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	57	5	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	57	6	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	57	7	advantage	advantage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	57	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	57	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	57	10	unique	unique	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	57	11	functional	functional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	57	12	nature	nature	NN
work_j4jyyr4gnrbelgfykpcumrgui4	57	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	57	14	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	57	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	58	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	58	2	therefore	therefore	RB
work_j4jyyr4gnrbelgfykpcumrgui4	58	3	in-	in-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	58	4	troduce	troduce	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	58	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	58	6	new	new	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	58	7	factorization	factorization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	58	8	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	58	9	call	call	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	58	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	58	11	Argument	Argument	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	58	12	Model	Model	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	58	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	59	1	It	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	59	2	exploits	exploit	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	59	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	59	4	fact	fact	NN
work_j4jyyr4gnrbelgfykpcumrgui4	59	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	59	6	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	59	7	imposes	impose	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	59	8	strong	strong	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	59	9	constraints	constraint	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	59	10	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	59	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	59	12	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	59	13	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	59	14	left	left	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	59	15	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	59	16	right	right	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	59	17	children	child	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	59	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	59	19	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	59	20	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	59	21	must	must	MD
work_j4jyyr4gnrbelgfykpcumrgui4	59	22	combine	combine	VB
work_j4jyyr4gnrbelgfykpcumrgui4	59	23	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	59	24	create	create	VB
work_j4jyyr4gnrbelgfykpcumrgui4	59	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	59	26	parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	59	27	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	59	28	via	via	IN
work_j4jyyr4gnrbelgfykpcumrgui4	59	29	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	59	30	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	59	31	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	59	32	combinators	combinator	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	59	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	60	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	60	2	practice	practice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	60	3	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	4	means	mean	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	60	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	60	6	given	give	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	60	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	8	parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	60	9	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	60	10	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	60	11	Z	Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	60	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	60	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	14	choice	choice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	60	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	60	16	combinator4	combinator4	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	60	17	c	c	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	60	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	60	19	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	20	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	60	21	Y	y	UH
work_j4jyyr4gnrbelgfykpcumrgui4	60	22	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	60	23	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	60	24	uniquely	uniquely	RB
work_j4jyyr4gnrbelgfykpcumrgui4	60	25	determine	determine	VB
work_j4jyyr4gnrbelgfykpcumrgui4	60	26	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	27	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	60	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	60	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	60	30	left	left	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	60	31	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	60	32	right	right	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	60	33	children	child	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	60	34	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	60	35	3Huang	3Huang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	60	36	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	60	37	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	60	38	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	61	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	2	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	61	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	4	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	61	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	61	6	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	7	deficient	deficient	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	61	8	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	9	variant	variant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	61	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	61	11	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	61	12	extension	extension	NN
work_j4jyyr4gnrbelgfykpcumrgui4	61	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	61	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	61	15	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	61	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	61	17	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	61	18	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	19	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	61	20	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	61	21	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	61	22	without	without	IN
work_j4jyyr4gnrbelgfykpcumrgui4	61	23	k	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	61	24	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	61	25	best	best	RB
work_j4jyyr4gnrbelgfykpcumrgui4	61	26	smoothing	smooth	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	61	27	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	61	28	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	61	29	underperform	underperform	VB
work_j4jyyr4gnrbelgfykpcumrgui4	61	30	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	61	31	published	publish	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	61	32	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	61	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	62	1	4If	4if	LS
work_j4jyyr4gnrbelgfykpcumrgui4	62	2	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	62	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	62	4	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	62	5	atomic	atomic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	62	6	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	62	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	62	8	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	62	9	application	application	NN
work_j4jyyr4gnrbelgfykpcumrgui4	62	10	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	62	11	possible	possible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	62	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	63	1	77	77	CD
work_j4jyyr4gnrbelgfykpcumrgui4	63	2	Parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	3	c	c	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	4	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	63	5	Left	Left	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	6	Right	Right	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	7	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	8	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	9	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	10	B0	b0	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	11	>	>	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	63	13	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	14	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	15	Z)/Y	z)/y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	16	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	17	B0	b0	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	18	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	19	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	20	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	63	21	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	22	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	23	Z)\Y	Z)\Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	24	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	25	>	>	FW
work_j4jyyr4gnrbelgfykpcumrgui4	63	26	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	27	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	28	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	29	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	30	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	31	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	32	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	33	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	34	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	35	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	36	Z	Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	37	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	38	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	63	39	correspondingly	correspondingly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	40	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	63	41	X\Z	x\z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	42	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	63	43	Parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	44	c	c	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	45	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	63	46	Left	Left	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	47	Right	Right	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	48	X\Z	x\z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	49	B0	b0	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	50	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	51	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	63	52	X\Z)/Y	X\Z)/Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	53	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	54	B0	b0	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	55	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	56	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	57	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	63	58	X\Z)\Y	X\Z)\Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	59	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	60	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	61	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	62	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	63	63	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	64	Y\Z	y\z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	65	B1	b1	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	66	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	63	67	Y\Z	Y\Z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	63	68	X\Y	x\y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	69	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	63	70	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	71	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	63	72	changing	change	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	63	73	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	63	74	raising	raising	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	75	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	63	76	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	77	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	63	78	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	63	79	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	63	80	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	81	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	63	82	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	83	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	63	84	treatment	treatment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	85	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	63	86	root	root	NN
work_j4jyyr4gnrbelgfykpcumrgui4	63	87	productions	production	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	63	88	extends	extend	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	63	89	easily	easily	RB
work_j4jyyr4gnrbelgfykpcumrgui4	63	90	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	63	91	handle	handle	VB
work_j4jyyr4gnrbelgfykpcumrgui4	63	92	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	63	93	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	63	94	unary	unary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	63	95	cases	case	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	63	96	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	64	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	64	2	simply	simply	RB
work_j4jyyr4gnrbelgfykpcumrgui4	64	3	treat	treat	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	64	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	5	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	6	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	7	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	9	unary	unary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	10	outcome	outcome	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	11	so	so	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	12	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	14	parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	16	combinator	combinator	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	64	18	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	19	uniquely	uniquely	RB
work_j4jyyr4gnrbelgfykpcumrgui4	64	20	specify	specify	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	64	21	every	every	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	22	detail	detail	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	25	unary	unary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	26	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	27	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	64	28	Parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	29	c	c	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	30	⇒	⇒	CD
work_j4jyyr4gnrbelgfykpcumrgui4	64	31	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	32	TOP	TOP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	33	TOP	TOP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	34	∈{S	∈{S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	35	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	36	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	37	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	38	S/(S\N	s/(s\n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	39	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	40	T	t	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	41	<	<	XX
work_j4jyyr4gnrbelgfykpcumrgui4	64	42	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	43	S\(S	S\(S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	44	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	64	45	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	46	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	47	T	t	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	48	>	>	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	49	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	50	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	64	51	still	still	RB
work_j4jyyr4gnrbelgfykpcumrgui4	64	52	distinguish	distinguish	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	64	53	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	54	same	same	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	55	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	56	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	64	57	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	58	before	before	RB
work_j4jyyr4gnrbelgfykpcumrgui4	64	59	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	60	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	61	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	62	unary	unary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	63	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	64	binary	binary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	64	65	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	66	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	67	left	leave	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	64	68	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	64	69	right	right	UH
work_j4jyyr4gnrbelgfykpcumrgui4	64	70	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	71	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	72	leading	lead	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	64	73	us	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	64	74	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	64	75	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	64	76	following	follow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	64	77	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	78	definition	definition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	79	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	64	80	Given	give	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	64	81	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	64	82	P	p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	83	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	64	84	=	=	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	64	85	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	86	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	64	87	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	88	where	where	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	64	89	type(t	type(t	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	90	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	91	∈{Left	∈{left	UH
work_j4jyyr4gnrbelgfykpcumrgui4	64	92	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	93	Right	right	UH
work_j4jyyr4gnrbelgfykpcumrgui4	64	94	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	95	Unary	Unary	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	96	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	97	Lex	Lex	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	98	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	99	p(t|P	p(t|p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	100	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	101	×	×	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	64	102	{	{	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	103	p(w|P	p(w|p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	104	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	105	t	t	LS
work_j4jyyr4gnrbelgfykpcumrgui4	64	106	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	107	Lex	Lex	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	108	p(Y|P	p(Y|P	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	109	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	110	t	t	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	111	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	112	×p(c|P	×p(c|p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	113	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	114	t	t	NN
work_j4jyyr4gnrbelgfykpcumrgui4	64	115	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	64	116	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	117	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	64	118	o.w	o.w	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	64	119	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	65	1	Argument	Argument	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	65	2	Combinator	Combinator	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	65	3	Note	note	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	65	4	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	65	5	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	65	6	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	65	7	generates	generate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	65	8	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	65	9	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	65	10	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	65	11	cat-	cat-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	65	12	egory	egory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	65	13	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	65	14	uniquely	uniquely	RB
work_j4jyyr4gnrbelgfykpcumrgui4	65	15	defines	define	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	65	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	65	17	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	65	18	children	child	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	65	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	65	20	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	65	21	par-	par-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	65	22	ent	ent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	65	23	node	node	RB
work_j4jyyr4gnrbelgfykpcumrgui4	65	24	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	66	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	66	2	will	will	MD
work_j4jyyr4gnrbelgfykpcumrgui4	66	3	see	see	VB
work_j4jyyr4gnrbelgfykpcumrgui4	66	4	below	below	IN
work_j4jyyr4gnrbelgfykpcumrgui4	66	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	66	6	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	66	7	greatly	greatly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	66	8	simpli-	simpli-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	66	9	fies	fie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	66	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	66	11	development	development	NN
work_j4jyyr4gnrbelgfykpcumrgui4	66	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	66	13	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	66	14	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	66	15	parametric	parametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	66	16	extensions	extension	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	66	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	67	1	5	5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	67	2	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	67	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	67	4	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	67	5	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	67	6	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	67	7	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	8	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	9	Simple	simple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	10	generative	generative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	11	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	67	12	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	13	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	67	14	PCFGs	PCFGs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	67	15	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	67	16	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	67	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	67	18	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	67	19	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	67	20	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	67	21	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	67	22	’s	’s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	23	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	67	24	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	25	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	67	26	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	67	27	robust	robust	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	28	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	67	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	67	30	face	face	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	31	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	67	32	sparsity	sparsity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	33	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	67	34	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	67	35	they	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	67	36	assign	assign	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	67	37	zero	zero	CD
work_j4jyyr4gnrbelgfykpcumrgui4	67	38	probability	probability	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	39	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	67	40	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	67	41	unseen	unseen	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	67	42	event	event	NN
work_j4jyyr4gnrbelgfykpcumrgui4	67	43	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	68	1	Sparsity	sparsity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	68	2	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	68	3	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	68	4	particular	particular	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	68	5	problem	problem	NN
work_j4jyyr4gnrbelgfykpcumrgui4	68	6	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	68	7	formalisms	formalism	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	68	8	like	like	IN
work_j4jyyr4gnrbelgfykpcumrgui4	68	9	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	68	10	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	68	11	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	68	12	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	68	13	rich	rich	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	68	14	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	68	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	68	16	object	object	NN
work_j4jyyr4gnrbelgfykpcumrgui4	68	17	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	68	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	69	1	Nonpara-	Nonpara-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	69	2	metric	metric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	69	3	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	69	4	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	69	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	69	6	e.g.	e.g.	RB
work_j4jyyr4gnrbelgfykpcumrgui4	70	1	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	2	Processes	Processes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	3	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	4	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	6	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	7	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	8	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	70	9	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	70	10	hierarchical	hierarchical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	70	11	variants	variant	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	70	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	13	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	14	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	70	15	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	16	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	18	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	19	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	70	21	generalizations	generalization	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	70	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	23	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	25	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	27	overcome	overcome	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	70	28	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	70	29	problem	problem	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	30	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	70	31	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	70	32	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	70	33	elegant	elegant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	70	34	manner	manner	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	35	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	36	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	70	37	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	70	38	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	70	39	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	70	40	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	70	41	state	state	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	42	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	70	43	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	70	44	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	70	45	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	70	46	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	70	47	art	art	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	48	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	49	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	70	50	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	70	51	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	52	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	53	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	54	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	55	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	56	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	57	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	58	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	70	59	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	60	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	70	61	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	62	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	63	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	64	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	70	65	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	66	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	70	67	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	70	68	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	70	69	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	70	70	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	70	71	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	71	1	They	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	71	2	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	71	3	im-	im-	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	71	4	pose	pose	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	71	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	71	6	rich	rich	RB
work_j4jyyr4gnrbelgfykpcumrgui4	71	7	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	71	8	getting	get	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	71	9	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	71	10	richer	rich	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	71	11	behavior	behavior	NN
work_j4jyyr4gnrbelgfykpcumrgui4	71	12	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	71	13	seems	seem	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	71	14	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	71	15	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	71	16	advantageous	advantageous	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	71	17	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	71	18	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	71	19	modeling	modeling	NN
work_j4jyyr4gnrbelgfykpcumrgui4	71	20	applications	application	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	71	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	72	1	By	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	72	2	contrast	contrast	NN
work_j4jyyr4gnrbelgfykpcumrgui4	72	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	72	4	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	72	5	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	72	6	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	72	7	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	72	8	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	72	9	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	72	10	propose	propose	VB
work_j4jyyr4gnrbelgfykpcumrgui4	72	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	72	12	weighted	weighted	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	72	13	top	top	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	72	14	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	72	15	k	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	72	16	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	72	17	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	72	18	address	address	VB
work_j4jyyr4gnrbelgfykpcumrgui4	72	19	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	72	20	issues	issue	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	72	21	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	72	22	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	72	23	ad	ad	FW
work_j4jyyr4gnrbelgfykpcumrgui4	72	24	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	72	25	hoc	hoc	FW
work_j4jyyr4gnrbelgfykpcumrgui4	72	26	manner	manner	NN
work_j4jyyr4gnrbelgfykpcumrgui4	72	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	73	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	73	2	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	73	3	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	73	4	introduced	introduce	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	73	5	above	above	IN
work_j4jyyr4gnrbelgfykpcumrgui4	73	6	lends	lend	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	73	7	it-	it-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	73	8	self	self	NN
work_j4jyyr4gnrbelgfykpcumrgui4	73	9	particularly	particularly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	73	10	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	73	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	73	12	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	73	13	extensions	extension	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	73	14	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	73	15	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	73	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	73	17	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	73	18	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	73	19	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	73	20	Pro-	Pro-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	73	21	cesses	cesse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	73	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	73	23	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	73	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	73	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	74	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	3	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	5	size	size	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	8	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	74	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	11	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	13	productions	production	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	74	14	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	74	15	fixed	fix	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	74	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	74	17	small	small	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	74	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	74	19	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	74	20	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	74	21	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	74	22	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	23	formulation	formulation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	24	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	25	infinite	infinite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	74	26	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	74	27	allow	allow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	74	28	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	29	easy	easy	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	74	30	extension	extension	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	31	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	74	32	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	74	33	future	future	NN
work_j4jyyr4gnrbelgfykpcumrgui4	74	34	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	75	1	Specifically	specifically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	75	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	75	3	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	75	4	frame-	frame-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	5	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	6	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	75	7	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	75	8	extensions	extension	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	75	9	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	75	10	grow	grow	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	75	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	75	12	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	13	during	during	IN
work_j4jyyr4gnrbelgfykpcumrgui4	75	14	parsing	parsing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	15	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	75	16	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	17	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	75	18	fully	fully	RB
work_j4jyyr4gnrbelgfykpcumrgui4	75	19	lexicalize	lexicalize	VB
work_j4jyyr4gnrbelgfykpcumrgui4	75	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	75	21	pro-	pro-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	75	22	ductions	duction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	75	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	76	1	Additionally	additionally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	76	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	76	3	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	4	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	76	5	current	current	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	6	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	7	uses	use	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	76	8	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	76	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	76	10	restricted	restrict	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	76	11	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	13	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	76	14	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	76	15	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	76	16	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	76	17	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	76	18	finite	finite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	19	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	21	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	76	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	76	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	76	24	literature	literature	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	25	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	76	26	generalized	generalized	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	27	variants	variant	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	76	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	29	composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	30	make	make	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	76	31	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	76	32	possible	possible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	33	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	34	gener-	gener-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	35	ate	ate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	36	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	76	37	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	76	38	unbounded	unbounded	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	76	39	arity	arity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	76	40	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	77	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	77	2	therefore	therefore	RB
work_j4jyyr4gnrbelgfykpcumrgui4	77	3	be-	be-	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	77	4	lieve	lieve	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	77	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	6	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	7	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	77	8	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	9	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	77	10	natural	natural	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	77	11	probabilistic	probabilistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	77	12	frame-	frame-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	13	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	14	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	15	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	77	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	77	17	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	18	HDPs	hdp	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	77	19	make	make	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	77	20	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	77	21	possible	possible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	77	22	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	23	con-	con-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	77	24	sider	sider	VB
work_j4jyyr4gnrbelgfykpcumrgui4	77	25	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	26	potentially	potentially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	77	27	infinite	infinite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	77	28	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	29	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	30	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	77	31	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	77	32	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	77	33	instantiate	instantiate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	77	34	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	35	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	77	36	slot	slot	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	37	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	77	38	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	39	allowing	allow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	77	40	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	41	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	42	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	77	43	capture	capture	VB
work_j4jyyr4gnrbelgfykpcumrgui4	77	44	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	45	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	77	46	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	77	47	preferences	preference	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	77	48	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	49	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	50	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	51	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	52	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	77	53	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	77	54	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	77	55	appear	appear	VB
work_j4jyyr4gnrbelgfykpcumrgui4	77	56	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	77	57	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	77	58	position	position	NN
work_j4jyyr4gnrbelgfykpcumrgui4	77	59	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	78	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	78	2	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	78	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	78	4	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	78	5	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	78	6	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	78	7	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	78	8	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	78	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	78	10	multinomials	multinomial	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	78	11	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	78	12	drawn	draw	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	78	13	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	78	14	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	78	15	corresponding	corresponding	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	78	16	n-	n-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	78	17	dimensional	dimensional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	78	18	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	78	19	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	78	20	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	79	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	79	2	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	79	3	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	79	4	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	79	5	DP	DP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	79	6	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	79	7	generalizes	generalize	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	79	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	79	9	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	79	10	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	79	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	79	12	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	79	13	infinite	infinite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	79	14	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	79	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	79	16	possible	possible	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	79	17	outcomes	outcome	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	79	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	79	19	allowing	allow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	79	20	us	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	79	21	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	79	22	deal	deal	VB
work_j4jyyr4gnrbelgfykpcumrgui4	79	23	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	79	24	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	79	25	potentially	potentially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	79	26	infinite	infinite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	79	27	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	79	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	79	29	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	79	30	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	79	31	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	79	32	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	80	1	DPs	dp	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	80	2	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	80	3	defined	define	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	80	4	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	80	5	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	80	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	80	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	80	8	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	9	dis-	dis-	IN
work_j4jyyr4gnrbelgfykpcumrgui4	80	10	tribution	tribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	11	H	h	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	12	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	80	13	corresponds	correspond	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	80	14	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	80	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	80	16	mean	mean	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	80	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	80	19	DP	DP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	80	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	80	21	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	80	22	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	80	23	concentration	concentration	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	24	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	80	25	shape	shape	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	26	parameter	parameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	80	27	α	α	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	80	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	81	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	2	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	81	3	Hi-	Hi-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	4	erarchical	erarchical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	81	5	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	6	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	7	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	81	8	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	9	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	81	10	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	11	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	81	13	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	81	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	81	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	81	16	there	there	EX
work_j4jyyr4gnrbelgfykpcumrgui4	81	17	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	81	18	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	81	19	hierarchy	hierarchy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	21	DPs	dp	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	81	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	81	23	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	81	24	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	81	26	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	27	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	29	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	81	30	DP	DP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	31	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	32	level	level	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	33	n	n	CC
work_j4jyyr4gnrbelgfykpcumrgui4	81	34	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	81	35	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	81	36	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	37	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	81	38	level	level	NN
work_j4jyyr4gnrbelgfykpcumrgui4	81	39	n−	n−	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	81	40	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	81	41	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	82	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	2	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	82	4	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	5	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	82	6	Figure	Figure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	7	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	82	8	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	82	9	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	82	10	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	11	reformulation	reformulation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	82	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	14	Argument	Argument	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	15	Model	Model	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	16	introduced	introduce	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	82	17	above	above	RB
work_j4jyyr4gnrbelgfykpcumrgui4	82	18	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	19	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	82	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	21	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	22	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	23	Processes.5	Processes.5	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	24	At	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	26	heart	heart	NN
work_j4jyyr4gnrbelgfykpcumrgui4	82	27	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	28	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	29	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	82	30	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	82	31	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	82	32	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	82	33	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	82	34	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	82	35	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	82	36	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	83	1	By	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	2	combining	combine	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	83	3	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	83	4	stick	stick	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	5	breaking	break	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	83	6	process	process	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	7	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	8	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	83	9	multino-	multino-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	83	10	mial	mial	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	11	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	12	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	83	13	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	83	14	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	83	15	define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	83	16	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	83	17	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	18	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	19	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	83	20	5An	5an	CD
work_j4jyyr4gnrbelgfykpcumrgui4	83	21	alternative	alternative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	83	22	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	83	23	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	24	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	25	semantic	semantic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	83	26	parsing	parsing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	83	27	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	28	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	83	29	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	83	30	proposed	propose	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	83	31	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	83	32	Kwiatkowski	Kwiatkowski	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	83	33	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	83	34	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	83	35	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	84	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	84	2	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	84	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	84	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	85	1	78	78	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	2	HDP	hdp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	85	4	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	5	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	6	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	7	Draw	draw	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	8	global	global	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	9	parameters	parameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	10	Define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	11	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	12	root	root	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	13	parameter	parameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	14	θTOP	θtop	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	15	Draw	draw	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	16	top	top	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	17	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	85	18	level	level	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	19	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	20	weights	weight	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	85	21	βY	βY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	22	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	23	GEM(αY	GEM(αY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	25	Draw	draw	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	26	top	top	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	27	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	85	28	level	level	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	29	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	30	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	31	βL	βL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	32	∼	∼	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	33	GEM(αL	GEM(αL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	34	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	35	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	36	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	37	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	38	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	39	z	z	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	40	∈{1	∈{1	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	41	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	42	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	43	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	44	...	...	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	45	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	46	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	47	Define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	48	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	49	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	50	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	51	parameters	parameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	52	θTz	θTz	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	53	Draw	Draw	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	54	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	55	parameters	parameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	56	φYz	φyz	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	85	57	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	58	DP(αY	dp(αy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	59	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	60	βY	βY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	61	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	62	Draw	draw	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	63	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	64	emission	emission	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	65	parameters	parameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	66	φLz	φLz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	67	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	68	DP(αL	DP(αL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	69	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	70	βL	βL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	71	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	72	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	73	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	74	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	75	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	76	y	y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	77	∈{1	∈{1	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	78	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	79	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	80	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	81	...	...	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	82	}	}	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	83	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	84	Define	Define	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	85	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	86	combinator	combinator	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	87	parameters	parameter	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	85	88	θCz	θCz	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	89	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	90	y	y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	91	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	92	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	93	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	94	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	95	parse	parse	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	96	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	97	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	98	Generate	Generate	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	99	root	root	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	100	node	node	RB
work_j4jyyr4gnrbelgfykpcumrgui4	85	101	zTOP	zTOP	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	85	102	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	103	Binomial(θTOP	Binomial(θTOP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	104	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	105	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	106	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	107	node	node	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	108	i	i	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	85	109	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	110	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	111	parse	parse	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	112	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	113	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	114	Choose	choose	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	115	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	116	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	117	ti	ti	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	118	∼	∼	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	119	Multinomial(θTzi	Multinomial(θTzi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	120	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	121	If	if	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	122	ti	ti	FW
work_j4jyyr4gnrbelgfykpcumrgui4	85	123	=	=	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	85	124	=	=	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	85	125	Lex	lex	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	126	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	127	Emit	emit	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	128	terminal	terminal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	129	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	130	xi	xi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	131	∼	∼	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	85	132	Multinomial(φLzi	Multinomial(φLzi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	133	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	134	If	if	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	135	ti	ti	FW
work_j4jyyr4gnrbelgfykpcumrgui4	85	136	=	=	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	85	137	=	=	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	85	138	Left	left	UH
work_j4jyyr4gnrbelgfykpcumrgui4	85	139	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	85	140	Right	right	UH
work_j4jyyr4gnrbelgfykpcumrgui4	85	141	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	85	142	Unary	Unary	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	143	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	85	144	Generate	Generate	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	145	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	146	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	147	yi	yi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	148	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	149	Multinomial(φYzi	Multinomial(φYzi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	150	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	151	Generate	Generate	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	152	combinator	combinator	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	153	ci	ci	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	154	∼	∼	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	155	Multinomial(θCzi	Multinomial(θCzi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	156	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	157	yi	yi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	158	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	159	Deterministically	deterministically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	85	160	create	create	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	85	161	zL(i	zL(i	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	162	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	163	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	164	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	85	165	zR(i	zr(i	FW
work_j4jyyr4gnrbelgfykpcumrgui4	85	166	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	167	if	if	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	168	binary	binary	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	169	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	170	zi	zi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	171	yi	yi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	172	ci	ci	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	173	zL(i	zL(i	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	174	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	175	zR(i	zR(i	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	176	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	177	xL(i	xL(i	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	178	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	179	xR(i	xR(i	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	180	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	181	z	z	XX
work_j4jyyr4gnrbelgfykpcumrgui4	85	182	∞	∞	XX
work_j4jyyr4gnrbelgfykpcumrgui4	85	183	∞y	∞y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	184	φY	φY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	185	θT	θt	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	186	θC	θc	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	187	φL	φl	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	188	βY	βY	.
work_j4jyyr4gnrbelgfykpcumrgui4	85	189	βL	βL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	190	Because	because	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	191	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	85	192	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	85	193	working	work	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	85	194	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	85	195	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	196	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	197	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	198	parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	199	zi	zi	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	200	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	85	201	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	85	202	yi	yi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	203	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	85	204	combinator	combinator	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	205	ci	ci	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	206	uniquely	uniquely	RB
work_j4jyyr4gnrbelgfykpcumrgui4	85	207	define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	85	208	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	85	209	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	85	210	children	child	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	211	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	85	212	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	213	zL(i),zR(i	zL(i),zR(i	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	85	214	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	215	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	85	216	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	86	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	86	2	dashed	dash	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	86	3	arrows	arrow	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	86	4	here	here	RB
work_j4jyyr4gnrbelgfykpcumrgui4	86	5	rep-	rep-	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	86	6	resent	resent	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	86	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	86	8	deterministic	deterministic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	86	9	process	process	NN
work_j4jyyr4gnrbelgfykpcumrgui4	86	10	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	86	11	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	86	12	generate	generate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	86	13	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	86	14	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	86	15	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	86	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	87	1	Figure	figure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	87	2	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	87	3	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	87	4	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	87	5	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	87	6	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	87	7	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	87	8	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	87	9	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	87	10	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	87	11	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	87	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	87	13	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	87	14	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	87	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	87	16	space	space	NN
work_j4jyyr4gnrbelgfykpcumrgui4	87	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	87	18	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	87	19	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	87	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	87	21	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	87	22	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	87	23	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	87	24	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	87	25	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	87	26	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	87	27	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	87	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	88	1	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	88	2	every	every	DT
work_j4jyyr4gnrbelgfykpcumrgui4	88	3	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	4	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	88	6	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	88	7	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	8	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	88	10	emission	emission	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	11	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	88	12	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	88	13	drawn	draw	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	88	14	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	88	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	88	16	corresponding	correspond	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	88	17	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	88	18	Processes	Processes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	88	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	89	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	89	2	addition	addition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	89	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	89	4	there	there	EX
work_j4jyyr4gnrbelgfykpcumrgui4	89	5	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	89	6	several	several	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	89	7	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	89	8	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	89	9	tied	tie	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	89	10	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	89	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	89	12	given	give	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	89	13	symbol	symbol	NN
work_j4jyyr4gnrbelgfykpcumrgui4	89	14	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	89	15	generating	generate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	89	16	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	89	17	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	89	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	89	19	combinators	combinator	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	89	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	89	21	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	89	22	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	89	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	90	1	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	90	2	whose	whose	WP$
work_j4jyyr4gnrbelgfykpcumrgui4	90	3	stick	stick	NN
work_j4jyyr4gnrbelgfykpcumrgui4	90	4	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	90	5	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	90	6	βY	βY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	90	7	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	90	8	correspond	correspond	VB
work_j4jyyr4gnrbelgfykpcumrgui4	90	9	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	90	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	90	11	frequency	frequency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	90	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	90	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	90	14	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	90	15	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	90	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	90	17	corpus	corpus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	90	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	91	1	Next	next	RB
work_j4jyyr4gnrbelgfykpcumrgui4	91	2	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	91	3	build	build	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	91	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	91	5	hierarchical	hierarchical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	91	6	component	component	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	91	8	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	91	9	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	10	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	91	11	choosing	choose	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	91	12	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	91	13	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	14	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	15	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	91	16	φY	φY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	91	17	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	91	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	91	19	again	again	RB
work_j4jyyr4gnrbelgfykpcumrgui4	91	20	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	91	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	91	22	space	space	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	91	24	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	91	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	91	26	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	91	27	every	every	DT
work_j4jyyr4gnrbelgfykpcumrgui4	91	28	parent	parent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	91	29	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	91	30	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	91	31	Z.	z.	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	92	2	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	3	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	4	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	92	5	drawn	draw	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	92	6	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	92	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	92	8	previously	previously	RB
work_j4jyyr4gnrbelgfykpcumrgui4	92	9	defined	define	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	92	10	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	11	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	92	13	allowing	allow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	92	14	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	92	15	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	92	16	important	important	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	92	17	level	level	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	18	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	92	19	parameter	parameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	20	tying	tie	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	92	21	across	across	IN
work_j4jyyr4gnrbelgfykpcumrgui4	92	22	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	92	23	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	92	24	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	92	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	93	1	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	93	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	93	3	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	4	DP	DP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	93	5	does	do	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	93	6	define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	93	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	93	8	mean	mean	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	9	around	around	IN
work_j4jyyr4gnrbelgfykpcumrgui4	93	10	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	93	11	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	93	12	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	13	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	93	14	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	93	15	drawn	draw	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	93	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	93	17	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	93	18	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	93	19	require	require	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	93	20	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	93	21	notion	notion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	22	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	93	23	variance	variance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	24	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	93	25	precision	precision	NN
work_j4jyyr4gnrbelgfykpcumrgui4	93	26	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	93	27	de-	de-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	93	28	termines	termine	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	93	29	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	93	30	similar	similar	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	93	31	individual	individual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	93	32	draws	draw	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	93	33	will	will	MD
work_j4jyyr4gnrbelgfykpcumrgui4	93	34	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	93	35	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	94	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	94	2	precision	precision	NN
work_j4jyyr4gnrbelgfykpcumrgui4	94	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	94	4	determined	determine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	94	5	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	94	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	94	7	magnitude	magnitude	NN
work_j4jyyr4gnrbelgfykpcumrgui4	94	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	94	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	94	10	hy-	hy-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	94	11	perparameter	perparameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	94	12	αY	αY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	94	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	95	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	95	2	hierarchy	hierarchy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	95	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	95	4	paralleled	parallel	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	95	5	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	95	6	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	95	7	productions	production	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	95	8	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	95	9	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	95	10	drawn	draw	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	95	11	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	95	12	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	95	13	unigram	unigram	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	95	14	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	95	15	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	95	16	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	95	17	terminal	terminal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	95	18	symbols	symbol	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	95	19	controlled	control	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	95	20	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	95	21	αL.	αL.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	96	1	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	96	2	simplicity	simplicity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	96	3	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	96	4	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	96	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	96	6	same	same	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	96	7	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	96	8	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	96	9	setting	set	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	96	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	96	11	values	value	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	96	12	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	96	13	αL	αL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	96	14	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	96	15	αY	αY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	96	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	97	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	97	2	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	97	3	experimen-	experimen-	FW
work_j4jyyr4gnrbelgfykpcumrgui4	97	4	tal	tal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	97	5	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	97	6	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	7	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	97	8	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	97	9	vary	vary	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	97	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	11	value	value	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	13	αY	αY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	97	14	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	15	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	16	function	function	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	19	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	20	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	21	outcomes	outcome	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	97	22	allowed	allow	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	97	23	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	25	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	26	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	27	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	28	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	97	29	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	97	30	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	31	corpus	corpus	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	32	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	33	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	97	34	case	case	NN
work_j4jyyr4gnrbelgfykpcumrgui4	97	35	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	97	36	terminal	terminal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	97	37	symbols	symbol	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	97	38	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	98	1	Specifically	specifically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	98	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	98	3	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	98	4	set	set	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	98	5	αY	αY	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	98	6	=	=	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	98	7	np	np	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	98	8	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	98	9	conditioning	condition	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	98	10	contexts	contexts	NN
work_j4jyyr4gnrbelgfykpcumrgui4	98	11	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	98	12	n	n	DT
work_j4jyyr4gnrbelgfykpcumrgui4	98	13	out-	out-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	98	14	comes	come	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	98	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	99	1	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	99	2	Liang	Liang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	99	3	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	99	4	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	99	5	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	100	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	100	2	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	100	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	100	4	found	find	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	100	5	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	100	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	100	7	ideal	ideal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	100	8	value	value	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	9	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	100	10	alpha	alpha	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	11	appears	appear	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	100	12	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	100	13	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	100	14	superlinear	superlinear	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	15	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	100	16	sub-	sub-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	100	17	quadratic	quadratic	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	18	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	100	19	n	n	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	100	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	100	21	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	100	22	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	100	23	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	100	24	where	where	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	100	25	p	p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	26	takes	take	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	100	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	100	28	values	value	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	100	29	0	0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	100	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	100	31	1.0	1.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	100	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	100	33	1.5	1.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	100	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	100	35	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	100	36	2.0	2.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	100	37	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	100	38	explore	explore	VB
work_j4jyyr4gnrbelgfykpcumrgui4	100	39	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	100	40	range	range	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	41	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	100	42	uniform	uniform	NN
work_j4jyyr4gnrbelgfykpcumrgui4	100	43	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	100	44	quadratic	quadratic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	100	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	101	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	101	2	setting	setting	NN
work_j4jyyr4gnrbelgfykpcumrgui4	101	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	101	4	α	α	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	101	5	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	101	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	101	7	only	only	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	101	8	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	101	9	parameter	parameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	101	10	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	101	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	101	12	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	101	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	102	1	By	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	102	2	controlling	control	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	102	3	preci-	preci-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	102	4	sion	sion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	102	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	102	6	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	102	7	tell	tell	VB
work_j4jyyr4gnrbelgfykpcumrgui4	102	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	102	9	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	102	10	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	102	11	what	what	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	102	12	extent	extent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	102	13	global	global	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	102	14	cor-	cor-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	102	15	pus	pus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	102	16	statistics	statistic	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	102	17	should	should	MD
work_j4jyyr4gnrbelgfykpcumrgui4	102	18	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	102	19	trusted	trust	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	102	20	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	103	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	103	2	believe	believe	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	103	3	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	103	4	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	103	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	103	6	similar	similar	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	103	7	effect	effect	NN
work_j4jyyr4gnrbelgfykpcumrgui4	103	8	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	103	9	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	103	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	103	11	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	103	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	103	13	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	103	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	103	15	’s	’	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	103	16	top	top	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	103	17	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	103	18	k	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	103	19	upweighting	upweighting	NN
work_j4jyyr4gnrbelgfykpcumrgui4	103	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	103	21	smoothing	smooth	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	103	22	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	103	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	104	1	One	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	104	2	advantage	advantage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	104	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	104	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	104	5	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	104	6	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	104	7	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	104	8	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	104	9	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	104	10	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	104	11	requires	require	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	104	12	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	104	13	single	single	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	104	14	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	104	15	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	104	16	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	104	17	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	104	18	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	104	19	binary	binary	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	104	20	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	104	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	105	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	2	contrast	contrast	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	3	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	4	similar	similar	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	5	proposals	proposal	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	6	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	7	CFGs	CFGs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	105	8	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	105	9	Liang	Liang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	10	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	11	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	12	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	14	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	105	15	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	105	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	17	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	105	18	impose	impose	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	105	19	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	20	formal	formal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	21	restrictions	restriction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	22	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	24	nonterminals	nonterminal	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	25	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	27	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	28	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	29	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	30	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	105	31	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	105	32	ap-	ap-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	105	33	pear	pear	VB
work_j4jyyr4gnrbelgfykpcumrgui4	105	34	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	35	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	36	rewrite	rewrite	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	37	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	38	X	X	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	39	→	→	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	105	40	Y	y	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	41	Z	z	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	42	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	43	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	44	greatly	greatly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	105	45	sim-	sim-	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	105	46	plifies	plifie	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	105	47	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	48	modeling	modeling	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	49	problem	problem	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	50	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	105	51	yielding	yield	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	105	52	effectively	effectively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	105	53	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	54	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	55	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	105	56	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	105	57	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	105	58	akin	akin	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	59	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	60	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	61	HMMs	hmm	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	62	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	105	63	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	105	64	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	65	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	105	66	avoids	avoid	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	105	67	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	105	68	need	need	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	69	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	105	70	capture	capture	VB
work_j4jyyr4gnrbelgfykpcumrgui4	105	71	correlations	correlation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	72	be-	be-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	105	73	79	79	CD
work_j4jyyr4gnrbelgfykpcumrgui4	105	74	tween	tween	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	75	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	76	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	105	77	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	78	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	79	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	80	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	105	81	Z.	Z.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	82	Variational	Variational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	83	Inference	Inference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	105	84	HDPs	HDPs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	105	85	need	need	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	105	86	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	105	87	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	105	88	estimated	estimate	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	105	89	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	105	90	approximate	approximate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	105	91	techniques	technique	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	105	92	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	106	1	As	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	106	2	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	3	alternative	alternative	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	4	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	106	5	Gibbs	Gibbs	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	106	6	sampling	sampling	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	7	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	106	8	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	9	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	106	10	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	11	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	13	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	106	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	106	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	16	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	106	17	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	106	18	exact	exact	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	20	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	106	21	typically	typically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	106	22	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	106	23	slow	slow	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	24	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	106	25	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	106	26	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	27	clear	clear	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	28	convergence	convergence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	29	criteria	criterion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	106	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	31	variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	32	inference	inference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	33	algorithms	algorithm	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	106	34	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	106	35	Bishop	Bishop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	36	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	37	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	106	38	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	106	39	Blei	Blei	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	40	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	106	41	Jordan	Jordan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	106	42	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	106	43	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	106	44	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	106	45	estimate	estimate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	106	46	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	47	parame-	parame-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	48	ters	ter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	106	49	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	106	50	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	51	truncated	truncate	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	106	52	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	53	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	106	54	maximize	maximize	VB
work_j4jyyr4gnrbelgfykpcumrgui4	106	55	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	56	lower	low	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	106	57	bound	bound	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	58	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	106	59	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	60	likelihood	likelihood	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	61	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	106	62	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	106	63	actual	actual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	106	64	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	106	65	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	107	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	107	2	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	107	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	107	4	factorization	factorization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	107	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	107	7	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	107	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	107	10	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	11	procedure	procedure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	12	analogous	analogous	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	107	13	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	107	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	107	15	Inside	Inside	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	107	16	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	107	17	Outside	outside	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	107	18	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	19	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	107	20	Lari	Lari	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	107	21	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	107	22	Young	Young	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	107	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	107	24	1991	1991	CD
work_j4jyyr4gnrbelgfykpcumrgui4	107	25	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	107	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	107	27	allowing	allow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	107	28	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	29	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	107	30	run	run	VB
work_j4jyyr4gnrbelgfykpcumrgui4	107	31	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	107	32	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	107	33	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	107	34	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	107	35	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	107	36	trivially	trivially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	107	37	parallelizable	parallelizable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	107	38	manner	manner	NN
work_j4jyyr4gnrbelgfykpcumrgui4	107	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	108	1	To	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	108	2	initialize	initialize	VB
work_j4jyyr4gnrbelgfykpcumrgui4	108	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	108	4	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	108	5	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	108	6	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	108	7	stick	stick	NN
work_j4jyyr4gnrbelgfykpcumrgui4	108	8	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	108	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	108	10	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	108	11	fol-	fol-	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	108	12	low	low	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	108	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	108	14	example	example	NN
work_j4jyyr4gnrbelgfykpcumrgui4	108	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	108	16	Kurihara	Kurihara	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	108	17	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	108	18	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	108	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	109	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	109	2	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	109	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	109	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	109	5	use	use	VB
work_j4jyyr4gnrbelgfykpcumrgui4	109	6	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	109	7	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	109	8	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	109	9	initialized	initialize	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	109	10	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	109	11	uniform	uniform	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	109	12	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	109	13	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	109	14	compute	compute	VB
work_j4jyyr4gnrbelgfykpcumrgui4	109	15	global	global	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	109	16	counts	count	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	109	17	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	109	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	109	19	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	109	20	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	109	21	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	109	22	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	109	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	110	1	When	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	110	2	normalized	normalize	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	110	3	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	110	4	provide	provide	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	110	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	110	6	better	well	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	110	7	initialization	initialization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	110	8	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	110	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	110	10	uniform	uniform	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	110	11	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	110	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	110	13	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	110	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	111	1	Updates	update	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	111	2	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	111	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	111	4	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	111	5	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	111	6	then	then	RB
work_j4jyyr4gnrbelgfykpcumrgui4	111	7	performed	perform	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	111	8	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	111	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	111	10	coordi-	coordi-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	11	nate	nate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	111	12	descent	descent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	13	manner	manner	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	14	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	111	15	includes	include	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	111	16	re	re	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	17	-	-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	18	estimation	estimation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	111	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	111	21	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	111	22	DPs	dp	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	111	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	112	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	112	2	variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	112	3	inference	inference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	112	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	112	5	multinomial	multinomial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	112	6	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	112	7	W	W	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	112	8	take	take	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	112	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	112	10	place	place	NN
work_j4jyyr4gnrbelgfykpcumrgui4	112	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	112	12	probabilities	probability	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	112	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	113	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	113	2	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	113	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	113	4	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	113	5	outcome	outcome	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	6	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	113	7	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	113	8	conditioning	conditioning	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	9	variable	variable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	113	10	P	p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	11	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	113	12	com-	com-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	113	13	puted	put	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	113	14	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	113	15	summing	sum	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	113	16	pseudocounts	pseudocount	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	113	17	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	113	18	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	113	19	scaled	scale	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	113	20	mean	mean	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	113	21	vector	vector	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	22	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	113	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	113	24	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	25	DP	dp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	113	26	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	114	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	2	computation	computation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	3	involves	involve	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	114	4	moving	move	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	114	5	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	7	direction	direction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	10	gradient	gradient	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	13	Dirich-	Dirich-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	14	let	let	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	114	15	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	16	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	114	17	results	result	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	114	18	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	20	following	follow	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	114	21	differ-	differ-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	114	22	ence	ence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	24	Digammas	Digammas	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	25	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	26	Ψ	Ψ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	27	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	28	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	114	29	WP	WP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	30	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	31	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	32	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	33	=	=	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	114	34	Ψ(C(P	ψ(c(p	XX
work_j4jyyr4gnrbelgfykpcumrgui4	114	35	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	114	36	Y	Y	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	37	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	38	+	+	CC
work_j4jyyr4gnrbelgfykpcumrgui4	114	39	α	α	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	40	P	p	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	41	βY	βy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	42	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	43	−	−	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	44	Ψ(C(P,∗	Ψ(C(P,∗	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	45	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	46	+	+	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	114	47	αP	αp	ADD
work_j4jyyr4gnrbelgfykpcumrgui4	114	48	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	114	49	Importantly	importantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	114	50	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	114	51	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	52	Digamma	Digamma	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	114	53	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	114	54	multinomial	multinomial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	114	55	weights	weight	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	114	56	comprise	comprise	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	114	57	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	58	righ	righ	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	59	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	114	60	get	get	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	61	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	114	62	richer	rich	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	114	63	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	64	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	114	65	biasing	bias	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	114	66	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	114	67	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	114	68	against	against	IN
work_j4jyyr4gnrbelgfykpcumrgui4	114	69	rare	rare	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	114	70	outcomes	outcome	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	114	71	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	115	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	115	2	addition	addition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	115	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	115	4	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	115	5	variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	115	6	inference	inference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	115	7	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	115	8	done	do	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	115	9	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	115	10	coordinate	coordinate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	115	11	descent	descent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	115	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	115	13	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	115	14	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	115	15	trivially	trivially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	115	16	parallelizeable	parallelizeable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	115	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	116	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	2	practice	practice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	116	4	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	5	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	116	6	testing	test	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	116	7	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	116	8	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	116	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	116	11	corpora	corpora	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	12	containing	contain	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	116	13	sen-	sen-	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	116	14	tences	tence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	116	15	up	up	RP
work_j4jyyr4gnrbelgfykpcumrgui4	116	16	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	116	17	length	length	VB
work_j4jyyr4gnrbelgfykpcumrgui4	116	18	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	116	19	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	116	20	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	21	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	116	22	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	23	takes	take	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	116	24	be-	be-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	116	25	tween	tween	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	26	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	116	27	minute	minute	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	28	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	29	at	at	RB
work_j4jyyr4gnrbelgfykpcumrgui4	116	30	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	116	31	three	three	CD
work_j4jyyr4gnrbelgfykpcumrgui4	116	32	hours	hour	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	116	33	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	34	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	116	35	single	single	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	116	36	12-core	12-core	CD
work_j4jyyr4gnrbelgfykpcumrgui4	116	37	machine	machine	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	38	depending	depend	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	116	39	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	116	40	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	116	41	size	size	NN
work_j4jyyr4gnrbelgfykpcumrgui4	116	42	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	117	1	6	6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	117	2	Evaluation	Evaluation	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	117	3	As	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	4	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	117	5	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	117	6	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	7	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	117	8	task	task	NN
work_j4jyyr4gnrbelgfykpcumrgui4	117	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	117	10	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	117	11	evaluate	evaluate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	117	12	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	117	13	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	117	14	against	against	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	15	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	117	16	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	117	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	18	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	117	19	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	117	20	treebanks	treebank	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	117	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	117	22	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	117	23	measure	measure	VB
work_j4jyyr4gnrbelgfykpcumrgui4	117	24	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	117	25	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	26	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	117	27	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	28	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	117	29	accuracy	accuracy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	117	30	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	117	31	directed	direct	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	117	32	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	117	33	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	117	34	i.e.	i.e.	FW
work_j4jyyr4gnrbelgfykpcumrgui4	118	1	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	118	2	percentage	percentage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	118	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	118	4	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	118	5	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	118	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	118	7	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	118	8	corpus	corpus	NN
work_j4jyyr4gnrbelgfykpcumrgui4	118	9	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	118	10	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	118	11	correctly	correctly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	118	12	attached	attach	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	118	13	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	118	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	119	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	119	2	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	119	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	119	4	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	119	5	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	119	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	119	7	PASCAL	PASCAL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	8	challenge	challenge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	119	9	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	119	10	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	119	11	induction	induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	13	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	119	14	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	119	15	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	16	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	119	18	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	119	19	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	119	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	119	22	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	119	23	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	119	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	119	25	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	26	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	119	27	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	28	shared	share	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	119	29	task	task	NN
work_j4jyyr4gnrbelgfykpcumrgui4	119	30	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	31	Buchholz	Buchholz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	32	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	119	33	Marsi	Marsi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	119	35	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	119	36	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	37	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	119	38	Goldberg	Goldberg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	119	39	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	40	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	119	41	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	119	42	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	119	43	Hebrew	Hebrew	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	119	44	corpus	corpus	NN
work_j4jyyr4gnrbelgfykpcumrgui4	119	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	120	1	Converting	convert	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	120	2	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	120	3	derivations	derivation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	120	4	into	into	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	5	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	120	6	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	120	7	mostly	mostly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	120	8	straightforward	straightforward	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	120	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	120	10	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	120	12	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	120	13	derivation	derivation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	14	identifies	identify	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	120	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	120	16	root	root	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	17	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	18	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	19	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	120	20	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	120	22	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	120	23	head-	head-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	120	24	argument	argument	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	120	26	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	27	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	120	28	modifier	modifi	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	120	29	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	120	30	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	120	31	easily	easily	RB
work_j4jyyr4gnrbelgfykpcumrgui4	120	32	read	read	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	120	33	off	off	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	34	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	35	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	120	36	derivations	derivation	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	120	37	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	120	38	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	120	39	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	120	40	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	120	41	de-	de-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	120	42	fines	fine	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	120	43	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	120	44	explicitly	explicitly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	120	45	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	121	1	Unlike	unlike	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	2	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	3	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	5	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	121	6	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	121	7	designed	design	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	121	8	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	121	9	recover	recover	VB
work_j4jyyr4gnrbelgfykpcumrgui4	121	10	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	11	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	12	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	13	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	14	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	121	15	arise	arise	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	121	16	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	17	control	control	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	121	19	binding	bind	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	121	20	constructions	construction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	21	as	as	RB
work_j4jyyr4gnrbelgfykpcumrgui4	121	22	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	121	23	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	24	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	25	wh	wh	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	26	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	121	27	extraction	extraction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	28	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	121	29	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	30	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	31	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	32	coordi-	coordi-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	33	nation	nation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	35	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	121	36	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	37	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	121	38	requires	require	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	121	39	re	re	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	40	-	-	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	41	entrancies	entrancie	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	42	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	43	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	121	44	co-	co-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	45	indexation	indexation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	46	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	47	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	48	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	121	49	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	121	50	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	121	51	Steed-	Steed-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	121	52	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	53	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	54	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	121	55	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	121	56	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	57	within	within	IN
work_j4jyyr4gnrbelgfykpcumrgui4	121	58	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	121	59	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	60	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	61	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	121	62	trigger	trigger	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	121	63	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	121	64	constructions	construction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	65	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	121	66	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	121	67	current	current	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	68	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	121	69	returns	return	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	121	70	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	121	71	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	121	72	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	121	73	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	122	1	But	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	122	2	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	122	3	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	122	4	gram-	gram-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	122	5	mars	mars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	122	6	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	122	7	captures	capture	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	122	8	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	122	9	local	local	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	122	10	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	122	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	122	12	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	122	13	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	122	14	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	122	15	negative	negative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	122	16	influence	influence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	122	17	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	122	18	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	122	19	current	current	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	122	20	evaluation	evaluation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	122	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	123	1	However	however	RB
work_j4jyyr4gnrbelgfykpcumrgui4	123	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	123	3	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	123	4	direct	direct	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	5	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	123	6	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	123	7	depen-	depen-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	123	8	dency	dency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	123	9	treebanks	treebank	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	123	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	123	11	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	123	12	produced	produce	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	123	13	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	123	14	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	123	15	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	123	16	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	123	17	difficult	difficult	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	18	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	123	19	Clark	Clark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	123	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	123	21	Curran	Curran	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	123	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	123	23	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	123	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	123	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	123	26	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	123	27	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	123	28	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	123	29	allow	allow	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	123	30	considerable	considerable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	31	freedom	freedom	NN
work_j4jyyr4gnrbelgfykpcumrgui4	123	32	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	123	33	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	123	34	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	123	35	analyze	analyze	VB
work_j4jyyr4gnrbelgfykpcumrgui4	123	36	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	37	constructions	construction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	123	38	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	39	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	123	40	verb	verb	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	123	41	clusters	cluster	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	123	42	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	123	43	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	123	44	verb	verb	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	123	45	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	123	46	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	123	47	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	123	48	?	?	.
work_j4jyyr4gnrbelgfykpcumrgui4	123	49	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	124	1	prepositional	prepositional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	124	2	phrases	phrase	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	124	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	124	4	particles	particle	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	124	5	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	124	6	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	124	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	124	8	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	124	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	124	10	noun	noun	NN
work_j4jyyr4gnrbelgfykpcumrgui4	124	11	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	124	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	124	13	preposition	preposition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	124	14	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	124	15	particle	particle	NN
work_j4jyyr4gnrbelgfykpcumrgui4	124	16	?	?	.
work_j4jyyr4gnrbelgfykpcumrgui4	125	1	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	125	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	125	3	subordinating	subordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	125	4	conjunctions	conjunction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	125	5	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	125	6	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	125	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	8	conjunction	conjunction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	9	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	10	dependent	dependent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	125	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	13	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	125	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	16	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	125	17	clause	clause	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	125	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	20	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	21	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	125	22	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	23	embed-	embed-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	125	24	ded	ded	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	125	25	clause	clause	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	26	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	27	dependent	dependent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	125	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	125	30	conjunction	conjunction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	125	31	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	125	32	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	125	33	vice	vice	RB
work_j4jyyr4gnrbelgfykpcumrgui4	125	34	versa	versa	RB
work_j4jyyr4gnrbelgfykpcumrgui4	125	35	?	?	.
work_j4jyyr4gnrbelgfykpcumrgui4	125	36	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	126	1	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	126	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	126	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	126	4	reflected	reflect	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	126	5	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	126	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	126	7	fact	fact	NN
work_j4jyyr4gnrbelgfykpcumrgui4	126	8	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	126	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	126	10	tree-	tree-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	126	11	banks	bank	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	126	12	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	126	13	consider	consider	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	126	14	often	often	RB
work_j4jyyr4gnrbelgfykpcumrgui4	126	15	apply	apply	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	126	16	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	126	17	conventions	convention	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	126	18	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	126	19	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	126	20	cases	case	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	126	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	127	1	Although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	127	2	remedying	remedy	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	127	3	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	4	issue	issue	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	5	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	127	6	beyond	beyond	IN
work_j4jyyr4gnrbelgfykpcumrgui4	127	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	8	scope	scope	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	127	10	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	11	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	127	13	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	14	discrepancies	discrepancy	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	127	15	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	127	16	much	much	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	127	17	hint	hint	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	18	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	127	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	20	need	need	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	21	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	127	22	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	127	23	better	well	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	127	24	mechanism	mechanism	NN
work_j4jyyr4gnrbelgfykpcumrgui4	127	25	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	127	26	evaluate	evaluate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	127	27	linguistically	linguistically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	127	28	equivalent	equivalent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	127	29	structures	structure	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	127	30	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	127	31	tree-	tree-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	127	32	bank	bank	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	127	33	standardization	standardization	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	127	34	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	128	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	128	2	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	128	3	problematic	problematic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	128	4	construction	construction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	128	5	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	128	6	coordina-	coordina-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	128	7	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	128	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	129	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	2	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	129	3	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	129	4	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	129	5	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	6	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	129	7	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	8	schemes	scheme	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	129	10	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	11	conjuncts	conjunct	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	12	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	129	13	independent	independent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	129	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	129	15	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	129	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	17	conjunction	conjunction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	18	itself	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	129	19	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	129	20	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	129	21	attached	attach	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	129	22	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	24	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	25	graph	graph	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	129	27	whereas	whereas	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	28	de-	de-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	129	29	pendency	pendency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	30	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	31	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	129	32	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	129	33	stipulate	stipulate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	129	34	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	35	either	either	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	36	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	129	37	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	38	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	39	conjuncts	conjunct	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	40	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	129	41	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	42	conjunction	conjunction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	43	itself	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	129	44	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	129	45	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	46	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	129	47	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	129	48	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	49	multiple	multiple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	129	50	possibilities	possibility	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	51	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	129	52	where	where	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	129	53	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	129	54	remaining	remain	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	129	55	80	80	CD
work_j4jyyr4gnrbelgfykpcumrgui4	129	56	constituents	constituent	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	129	57	attach	attach	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	129	58	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	130	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	2	addition	addition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	3	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	130	5	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	6	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	130	7	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	130	9	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	130	10	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	130	11	identified	identify	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	130	12	five	five	CD
work_j4jyyr4gnrbelgfykpcumrgui4	130	13	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	14	styles	style	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	130	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	16	con-	con-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	17	junction	junction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	18	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	19	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	130	20	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	130	21	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	130	22	Figure	figure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	23	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	130	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	130	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	130	26	although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	27	several	several	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	28	cor-	cor-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	29	pora	pora	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	130	30	distinguish	distinguish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	31	multiple	multiple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	32	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	130	33	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	130	34	coordinating	coordinate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	130	35	con-	con-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	130	36	junctions	junction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	130	37	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	130	38	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	130	39	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	130	40	styles	style	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	130	41	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	130	42	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	130	43	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	130	44	shown	show	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	130	45	here	here	RB
work_j4jyyr4gnrbelgfykpcumrgui4	130	46	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	130	47	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	131	1	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	131	2	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	131	3	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	4	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	131	5	explicit	explicit	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	131	6	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	131	7	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	131	8	coordi-	coordi-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	9	nation	nation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	131	11	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	131	12	transform	transform	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	131	13	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	131	14	output	output	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	15	into	into	IN
work_j4jyyr4gnrbelgfykpcumrgui4	131	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	131	17	desired	desire	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	131	18	target	target	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	19	representation	representation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	20	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	131	21	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	131	22	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	131	23	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	131	24	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	131	25	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	131	26	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	132	1	7	7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	132	2	Experiments	experiment	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	132	3	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	132	4	evaluate	evaluate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	132	5	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	132	6	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	132	7	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	132	8	13	13	CD
work_j4jyyr4gnrbelgfykpcumrgui4	132	9	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	132	10	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	132	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	133	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	133	2	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	133	3	case	case	NN
work_j4jyyr4gnrbelgfykpcumrgui4	133	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	133	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	133	6	follow	follow	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	133	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	133	8	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	133	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	133	10	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	133	11	regimes	regime	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	133	12	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	133	13	were	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	133	14	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	133	15	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	133	16	obtain	obtain	VB
work_j4jyyr4gnrbelgfykpcumrgui4	133	17	previously	previously	RB
work_j4jyyr4gnrbelgfykpcumrgui4	133	18	published	publish	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	133	19	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	133	20	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	133	21	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	133	22	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	133	23	allow	allow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	133	24	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	133	25	direct	direct	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	133	26	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	133	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	134	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	134	2	com-	com-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	134	3	pare	pare	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	134	4	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	134	5	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	134	6	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	134	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	134	8	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	134	9	presented	present	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	134	10	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	134	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	134	12	PAS-	PAS-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	13	CAL	CAL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	14	Challenge	Challenge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	15	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	134	16	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	17	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	18	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	134	19	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	134	20	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	134	21	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	22	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	134	24	2012)6	2012)6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	134	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	134	26	as	as	RB
work_j4jyyr4gnrbelgfykpcumrgui4	134	27	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	134	28	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	134	29	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	134	30	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	31	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	32	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	134	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	135	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	135	2	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	135	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	135	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	135	5	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	135	6	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	135	7	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	135	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	136	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	136	2	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	136	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	136	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	137	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	137	2	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	137	3	Nivre	Nivre	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	4	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	5	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	137	6	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	7	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	137	8	Penn2Malt	penn2malt	NN
work_j4jyyr4gnrbelgfykpcumrgui4	137	9	implementation	implementation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	137	10	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	137	11	Collins	Collins	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	13	2003	2003	CD
work_j4jyyr4gnrbelgfykpcumrgui4	137	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	15	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	137	16	head	head	NN
work_j4jyyr4gnrbelgfykpcumrgui4	137	17	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	137	18	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	137	19	translate	translate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	137	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	137	21	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	22	Penn	Penn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	23	Treebank	Treebank	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	24	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	25	Marcus	Marcus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	26	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	27	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	28	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	137	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	137	30	1993	1993	CD
work_j4jyyr4gnrbelgfykpcumrgui4	137	31	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	137	32	into	into	IN
work_j4jyyr4gnrbelgfykpcumrgui4	137	33	dependencies	dependency	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	137	34	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	138	1	Finally	finally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	138	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	138	3	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	138	4	train-	train-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	5	ing	e	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	138	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	138	7	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	138	8	version	version	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	138	10	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	138	11	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	12	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	138	13	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	138	14	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	138	15	simple	simple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	138	16	smoothing	smoothing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	17	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	18	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	138	19	defines	define	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	138	20	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	138	21	small	small	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	138	22	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	23	proba-	proba-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	24	bility	bility	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	25	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	138	26	e−15	e−15	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	138	27	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	138	28	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	138	29	prevent	prevent	VB
work_j4jyyr4gnrbelgfykpcumrgui4	138	30	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	138	31	rule	rule	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	32	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	138	33	during	during	IN
work_j4jyyr4gnrbelgfykpcumrgui4	138	34	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	138	35	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	138	36	going	go	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	138	37	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	138	38	zero	zero	CD
work_j4jyyr4gnrbelgfykpcumrgui4	138	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	139	1	7.1	7.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	139	2	PASCAL	pascal	NN
work_j4jyyr4gnrbelgfykpcumrgui4	139	3	Challenge	Challenge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	4	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	5	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	6	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	7	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	8	Table	Table	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	9	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	139	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	139	11	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	139	12	compare	compare	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	139	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	139	14	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	139	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	139	17	ba-	ba-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	18	sic	sic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	19	Argument	Argument	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	20	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	139	21	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	22	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	23	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	139	25	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	26	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	139	27	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	28	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	139	29	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	30	four	four	CD
work_j4jyyr4gnrbelgfykpcumrgui4	139	31	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	139	32	settings	setting	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	139	33	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	34	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	139	35	hyperparameters	hyperparameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	139	36	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	37	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	38	ex-	ex-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	139	39	plained	plain	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	139	40	above	above	RB
work_j4jyyr4gnrbelgfykpcumrgui4	139	41	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	42	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	139	43	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	44	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	139	45	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	139	46	presented	present	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	139	47	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	48	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	139	49	PASCAL	PASCAL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	50	Challenge	Challenge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	51	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	139	52	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	53	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	54	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	55	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	139	56	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	139	57	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	58	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	139	59	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	139	60	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	139	61	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	139	62	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	140	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	140	2	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	140	3	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	140	4	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	140	5	competition	competition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	6	were	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	140	7	instructed	instruct	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	140	8	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	140	9	train	train	VB
work_j4jyyr4gnrbelgfykpcumrgui4	140	10	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	140	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	140	12	full	full	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	140	13	dataset	dataset	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	140	15	including	include	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	140	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	140	17	unlabelled	unlabelled	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	140	18	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	19	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	140	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	140	21	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	140	22	include	include	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	140	23	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	24	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	140	25	Hocken-	Hocken-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	26	maier	maier	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	27	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	140	28	2012a	2012a	CD
work_j4jyyr4gnrbelgfykpcumrgui4	140	29	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	140	30	’s	’s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	31	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	32	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	140	33	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	140	34	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	140	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	140	36	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	37	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	140	38	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	140	39	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	40	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	41	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	140	42	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	141	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	141	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	4	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	141	5	reimplementation	reimplementation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	141	7	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	141	9	Manning	Manning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	11	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	141	12	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	13	’s	’s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	14	DMV	DMV	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	15	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	16	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	141	17	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	141	18	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	19	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	141	20	substitution	substitution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	21	gram-	gram-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	22	mar	mar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	23	framework	framework	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	24	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	25	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	141	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	141	28	as	as	RB
work_j4jyyr4gnrbelgfykpcumrgui4	141	29	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	141	30	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	141	31	three	three	CD
work_j4jyyr4gnrbelgfykpcumrgui4	141	32	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	141	33	de-	de-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	34	pendency	pendency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	141	35	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	141	36	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	141	37	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	141	38	either	either	DT
work_j4jyyr4gnrbelgfykpcumrgui4	141	39	incorporate	incorporate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	141	40	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	41	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	42	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	141	43	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	142	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	142	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	142	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	142	4	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	142	5	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	142	6	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	142	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	142	8	deterministic	deterministic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	142	9	fash-	fash-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	142	10	ion	ion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	142	11	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	142	12	Søgaard	Søgaard	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	142	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	142	14	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	142	15	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	142	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	142	17	rely	rely	VB
work_j4jyyr4gnrbelgfykpcumrgui4	142	18	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	142	19	extensive	extensive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	142	20	tuning	tuning	NN
work_j4jyyr4gnrbelgfykpcumrgui4	142	21	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	142	22	6Numbers	6number	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	142	23	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	142	24	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	142	25	personal	personal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	142	26	correspondence	correspondence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	142	27	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	142	28	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	142	29	orga-	orga-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	142	30	nizers	nizer	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	142	31	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	143	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	143	2	previously	previously	RB
work_j4jyyr4gnrbelgfykpcumrgui4	143	3	published	publish	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	143	4	numbers	number	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	143	5	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	143	6	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	143	7	comparable	comparable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	143	8	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	143	9	literature	literature	VB
work_j4jyyr4gnrbelgfykpcumrgui4	143	10	due	due	IN
work_j4jyyr4gnrbelgfykpcumrgui4	143	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	143	12	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	143	13	error	error	NN
work_j4jyyr4gnrbelgfykpcumrgui4	143	14	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	143	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	143	16	evaluation	evaluation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	143	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	144	1	http://wiki	http://wiki	UH
work_j4jyyr4gnrbelgfykpcumrgui4	144	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	145	1	cs.ox.ac.uk/InducingLinguisticStructure/	cs.ox.ac.uk/inducinglinguisticstructure/	NN
work_j4jyyr4gnrbelgfykpcumrgui4	145	2	ResultsDepComparable	ResultsDepComparable	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	145	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	145	4	development	development	NN
work_j4jyyr4gnrbelgfykpcumrgui4	145	5	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	145	6	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	145	7	Tu	Tu	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	145	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	145	9	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	145	10	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	145	11	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	145	12	incorporate	incorporate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	145	13	mil-	mil-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	145	14	lions	lion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	145	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	145	16	additional	additional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	145	17	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	145	18	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	145	19	Wikipedia	Wikipedia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	145	20	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	145	21	esti-	esti-	VB
work_j4jyyr4gnrbelgfykpcumrgui4	145	22	mate	mate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	145	23	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	145	24	parameters	parameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	145	25	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	145	26	Marecek	Marecek	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	145	27	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	145	28	Zabokrtsky	Zabokrtsky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	145	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	145	30	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	145	31	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	145	32	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	146	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	146	2	ignore	ignore	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	146	3	punctuation	punctuation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	4	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	146	5	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	146	6	experiments	experiment	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	146	7	reported	report	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	146	8	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	146	9	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	146	10	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	146	12	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	146	13	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	146	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	146	15	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	16	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	146	17	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	146	18	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	146	19	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	146	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	146	21	evaluation	evaluation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	22	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	146	23	includes	include	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	146	24	punctuation	punctuation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	25	marks	mark	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	146	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	146	27	participants	participant	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	146	28	were	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	146	29	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	146	30	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	146	31	choose	choose	VB
work_j4jyyr4gnrbelgfykpcumrgui4	146	32	whether	whether	IN
work_j4jyyr4gnrbelgfykpcumrgui4	146	33	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	146	34	include	include	VB
work_j4jyyr4gnrbelgfykpcumrgui4	146	35	punctuation	punctuation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	146	36	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	146	37	ignore	ignore	VB
work_j4jyyr4gnrbelgfykpcumrgui4	146	38	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	146	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	147	1	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	147	2	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	147	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	147	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	147	5	only	only	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	147	6	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	147	7	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	8	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	147	9	directly	directly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	147	10	interpretable	interpretable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	147	11	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	147	12	output	output	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	147	14	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	147	15	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	147	16	include	include	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	147	17	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	147	18	di-	di-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	19	rect	rect	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	20	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	21	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	147	22	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	147	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	147	24	whose	whose	WP$
work_j4jyyr4gnrbelgfykpcumrgui4	147	25	TSG	TSG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	147	26	representa-	representa-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	27	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	147	28	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	147	29	equally	equally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	147	30	expressive	expressive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	147	31	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	147	32	ours	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	147	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	148	1	Finally	finally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	148	2	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	148	3	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	148	4	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	148	5	row	row	NN
work_j4jyyr4gnrbelgfykpcumrgui4	148	6	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	148	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	148	8	maximum	maximum	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	148	9	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	148	10	among	among	IN
work_j4jyyr4gnrbelgfykpcumrgui4	148	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	148	12	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	148	13	three	three	CD
work_j4jyyr4gnrbelgfykpcumrgui4	148	14	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	148	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	149	1	As	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	2	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	149	3	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	149	4	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	149	5	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	149	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	7	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	149	8	much	much	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	149	9	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	10	was	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	149	11	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	149	12	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	149	14	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	149	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	16	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	149	17	sys-	sys-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	149	18	tems	tem	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	19	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	149	20	simply	simply	RB
work_j4jyyr4gnrbelgfykpcumrgui4	149	21	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	149	22	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	23	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	24	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	25	trained	train	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	149	26	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	27	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	149	28	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	149	29	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	149	30	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	149	31	including	include	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	149	32	punctuation	punctuation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	149	33	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	149	34	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	35	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	149	36	then	then	RB
work_j4jyyr4gnrbelgfykpcumrgui4	149	37	evaluated	evaluate	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	149	38	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	149	39	lengths	length	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	149	40	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	149	41	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	149	42	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	149	43	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	150	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	150	2	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	150	3	version	version	NN
work_j4jyyr4gnrbelgfykpcumrgui4	150	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	150	5	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	150	6	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	150	7	shows	show	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	150	8	rather	rather	RB
work_j4jyyr4gnrbelgfykpcumrgui4	150	9	vari-	vari-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	150	10	able	able	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	150	11	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	150	12	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	150	13	although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	150	14	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	150	15	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	150	16	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	150	17	particu-	particu-	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	150	18	larly	larly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	150	19	bad	bad	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	150	20	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	150	21	Basque	Basque	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	150	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	150	23	Eu	Eu	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	150	24	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	150	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	150	26	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	150	27	outperforms	outperform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	150	28	both	both	CC
work_j4jyyr4gnrbelgfykpcumrgui4	150	29	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	150	30	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	150	31	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	150	32	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	150	33	some	some	DT
work_j4jyyr4gnrbelgfykpcumrgui4	150	34	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	150	35	settings	setting	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	150	36	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	151	1	By	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	151	2	contrast	contrast	NN
work_j4jyyr4gnrbelgfykpcumrgui4	151	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	151	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	151	5	HDP	hdp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	151	6	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	151	7	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	151	8	always	always	RB
work_j4jyyr4gnrbelgfykpcumrgui4	151	9	better	well	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	151	10	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	151	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	151	12	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	151	13	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	151	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	152	1	It	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	152	2	outperforms	outperform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	152	3	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	152	4	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	152	5	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	152	6	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	152	7	half	half	NN
work_j4jyyr4gnrbelgfykpcumrgui4	152	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	152	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	152	10	cor-	cor-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	152	11	pora	pora	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	152	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	153	1	On	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	153	2	average	average	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	153	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	153	4	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	153	5	outperforms	outperform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	153	6	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	153	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	153	8	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	153	9	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	153	10	10.3	10.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	11	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	153	13	9.3	9.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	14	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	15	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	153	16	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	17	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	153	19	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	153	20	9.7	9.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	21	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	22	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	153	23	7.8	7.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	24	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	25	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	153	26	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	153	27	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	153	28	respectively	respectively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	153	29	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	154	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	154	2	main	main	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	154	3	reason	reason	NN
work_j4jyyr4gnrbelgfykpcumrgui4	154	4	why	why	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	154	5	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	154	6	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	154	7	does	do	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	154	8	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	154	9	outperform	outperform	VB
work_j4jyyr4gnrbelgfykpcumrgui4	154	10	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	154	11	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	154	12	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	154	13	even	even	RB
work_j4jyyr4gnrbelgfykpcumrgui4	154	14	higher	high	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	154	15	margin	margin	NN
work_j4jyyr4gnrbelgfykpcumrgui4	154	16	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	154	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	154	18	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	154	19	obvious	obvious	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	154	20	11.4%/11.5	11.4%/11.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	154	21	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	154	22	deficit	deficit	NN
work_j4jyyr4gnrbelgfykpcumrgui4	154	23	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	154	24	Slovene	Slovene	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	154	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	155	1	However	however	RB
work_j4jyyr4gnrbelgfykpcumrgui4	155	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	155	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	155	4	Slovene	Slovene	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	155	5	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	155	6	tree-	tree-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	155	7	bank	bank	NN
work_j4jyyr4gnrbelgfykpcumrgui4	155	8	seems	seem	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	155	9	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	155	10	follow	follow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	155	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	155	12	substantially	substantially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	155	13	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	155	14	anno-	anno-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	155	15	tation	tation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	155	16	scheme	scheme	NN
work_j4jyyr4gnrbelgfykpcumrgui4	155	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	156	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	2	particular	particular	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	5	gold	gold	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	6	standard	standard	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	7	an-	an-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	156	8	notation	notation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	11	1,000	1,000	CD
work_j4jyyr4gnrbelgfykpcumrgui4	156	12	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	13	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	15	Slovene	Slovene	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	16	de-	de-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	17	velopment	velopment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	18	set	set	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	156	19	treats	treat	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	20	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	21	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	22	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	156	23	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	24	consisting	consist	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	156	25	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	26	independent	independent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	27	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	28	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	156	29	often	often	RB
work_j4jyyr4gnrbelgfykpcumrgui4	156	30	separated	separate	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	156	31	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	32	punctua-	punctua-	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	33	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	34	marks	mark	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	35	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	36	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	156	37	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	38	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	39	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	40	access	access	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	41	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	42	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	156	43	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	44	so	so	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	45	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	46	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	47	average	average	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	48	number	number	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	49	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	50	roots	root	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	51	per	per	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	52	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	53	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	54	2.7	2.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	156	55	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	156	56	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	156	57	>	>	XX
work_j4jyyr4gnrbelgfykpcumrgui4	156	58	“	"	``
work_j4jyyr4gnrbelgfykpcumrgui4	156	59	verjeti	verjeti	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	60	believe	believe	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	156	61	ti	ti	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	62	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	63	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	64	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	65	〈	〈	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	66	〈	〈	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	67	”	"	''
work_j4jyyr4gnrbelgfykpcumrgui4	156	68	je	je	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	69	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	70	mehko	mehko	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	71	soft	soft	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	72	rekla	rekla	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	73	said	say	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	156	74	When	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	156	75	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	156	76	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	77	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	78	presented	present	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	156	79	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	80	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	81	short	short	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	82	components	component	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	83	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	84	isolation	isolation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	85	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	86	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	156	87	oftentimes	oftentimes	RB
work_j4jyyr4gnrbelgfykpcumrgui4	156	88	analyzes	analyze	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	89	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	156	90	correctly	correctly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	156	91	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	92	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	156	93	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	94	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	156	95	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	156	96	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	156	97	return	return	VB
work_j4jyyr4gnrbelgfykpcumrgui4	156	98	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	99	tree	tree	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	100	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	156	101	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	156	102	sin-	sin-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	156	103	gle	gle	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	156	104	root	root	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	105	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	156	106	its	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	156	107	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	156	108	degrades	degrade	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	156	109	substantially	substantially	RB
work_j4jyyr4gnrbelgfykpcumrgui4	156	110	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	157	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	157	2	believe	believe	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	157	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	4	HDP	hdp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	5	performs	perform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	157	6	so	so	RB
work_j4jyyr4gnrbelgfykpcumrgui4	157	7	well	well	RB
work_j4jyyr4gnrbelgfykpcumrgui4	157	8	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	9	com-	com-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	10	pared	pare	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	157	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	13	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	14	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	15	because	because	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	16	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	18	influence	influence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	21	shared	shared	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	157	22	base	base	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	23	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	25	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	157	26	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	157	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	28	81	81	CD
work_j4jyyr4gnrbelgfykpcumrgui4	157	29	Ar	Ar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	31	Eu	Eu	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	33	Cs	Cs	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	35	Nl	Nl	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	36	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	37	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	38	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	39	Ch	Ch	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	40	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	41	He	He	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	42	Da	Da	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	43	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	44	He	He	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	45	Es	Es	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	46	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	47	Bg	Bg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	48	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	49	De	De	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	50	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	51	Pt	Pt	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	52	Sv	Sv	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	53	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	157	54	Sl	Sl	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	55	Ja	Ja	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	56	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	57	conj	conj	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	58	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	59	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	60	conj	conj	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	61	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	62	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	63	conj	conj	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	64	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	65	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	66	conj	conj	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	67	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	68	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	69	conj	conj	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	70	noun	noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	71	Figure	Figure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	157	72	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	157	73	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	157	74	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	75	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	157	76	treebanks	treebank	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	157	77	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	157	78	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	79	evaluation	evaluation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	80	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	157	81	standards	standard	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	157	82	exist	exist	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	157	83	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	157	84	annotating	annotate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	157	85	coordination	coordination	NN
work_j4jyyr4gnrbelgfykpcumrgui4	157	86	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	158	1	While	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	158	2	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	158	3	exhaustive	exhaustive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	158	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	158	5	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	158	6	table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	158	7	demonstrates	demonstrate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	158	8	five	five	CD
work_j4jyyr4gnrbelgfykpcumrgui4	158	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	158	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	158	11	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	158	12	common	common	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	158	13	schemes	scheme	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	158	14	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	158	15	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	158	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	158	17	literature	literature	NN
work_j4jyyr4gnrbelgfykpcumrgui4	158	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	159	1	Syntactically	syntactically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	159	2	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	159	3	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	159	4	identical	identical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	159	5	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	159	6	traditionally	traditionally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	159	7	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	159	8	draws	draw	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	159	9	arcs	arcs	RB
work_j4jyyr4gnrbelgfykpcumrgui4	159	10	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	159	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	159	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	159	13	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	159	14	without	without	IN
work_j4jyyr4gnrbelgfykpcumrgui4	159	15	attaching	attach	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	159	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	159	17	conjunction	conjunction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	159	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	160	1	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	160	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	160	3	purposes	purpose	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	160	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	160	5	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	160	6	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	160	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	160	8	literature	literature	NN
work_j4jyyr4gnrbelgfykpcumrgui4	160	9	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	160	10	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	160	11	implemented	implement	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	160	12	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	160	13	five	five	CD
work_j4jyyr4gnrbelgfykpcumrgui4	160	14	translation	translation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	160	15	schemes	scheme	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	160	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	161	1	Arabic	Arabic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	2	Danish	danish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	3	Slovene	slovene	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	4	Swedish	swedish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	5	Dutch	dutch	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	6	Basque	basque	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	7	Portuguese	portuguese	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	8	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	9	Childes	childe	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	161	10	Czech	czech	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	11	#	#	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	12	Tokens	token	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	161	13	5,470	5,470	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	14	25,341	25,341	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	15	54,032	54,032	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	16	61,877	61,877	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	17	78,737	78,737	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	18	81,345	81,345	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	19	158,648	158,648	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	20	163,727	163,727	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	21	290,604	290,604	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	22	436,126	436,126	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	23	#	#	$
work_j4jyyr4gnrbelgfykpcumrgui4	161	24	Tags	Tags	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	25	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	26	24	24	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	27	36	36	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	28	30	30	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	29	304	304	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	30	14	14	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	31	23	23	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	32	36	36	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	33	69	69	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	34	62	62	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	35	PA	pa	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	36	SC	SC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	37	A	A	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	38	L	L	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	39	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	40	60.8/58.4	60.8/58.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	41	44.7/39.4	44.7/39.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	42	62.6/57.9	62.6/57.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	43	63.2/56.6	63.2/56.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	44	51.8/52.0	51.8/52.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	45	53.0/48.9	53.0/48.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	46	52.4/50.2	52.4/50.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	47	68.6/63.3	68.6/63.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	48	47.4/46.1	47.4/46.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	49	47.9/43.1	47.9/43.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	50	Max	Max	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	51	67.2/66.8	67.2/66.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	52	60.1/56.0	60.1/56.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	53	65.6/61.8	65.6/61.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	54	72.8/63.4	72.8/63.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	55	51.1/47.6	51.1/47.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	56	53.7/47.8	53.7/47.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	57	67.0/61.8	67.0/61.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	58	71.2/64.8	71.2/64.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	59	56.0/54.5	56.0/54.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	60	58.3/54.4	58.3/54.4	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	61	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	62	41.6/43.7	41.6/43.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	63	46.4/43.8	46.4/43.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	64	49.6/43.9	49.6/43.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	65	63.7/57.0	63.7/57.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	66	49.7/43.6	49.7/43.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	67	45.1/39.6	45.1/39.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	68	70.8/67.2	70.8/67.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	69	68.2/59.6	68.2/59.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	70	61.4/59.8	61.4/59.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	71	45.0/38.9	45.0/38.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	72	T	T	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	73	hi	hi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	74	s	s	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	75	w	w	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	76	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	161	77	k	k	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	78	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	79	41.6/42.9	41.6/42.9	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	80	43.4/39.2	43.4/39.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	81	46.1/41.1	46.1/41.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	82	70.1/59.7	70.1/59.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	83	52.2/47.2	52.2/47.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	84	29.6/26.5	29.6/26.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	85	62.2/59.7	62.2/59.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	86	59.5/52.4	59.5/52.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	87	53.3/51.9	53.3/51.9	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	88	50.5/45.8	50.5/45.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	89	HDP0.0	HDP0.0	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	90	48.0/50.0	48.0/50.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	91	63.9/58.5	63.9/58.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	92	44.8/39.8	44.8/39.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	93	67.6/62.1	67.6/62.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	94	45.0/33.9	45.0/33.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	95	41.6/39.1	41.6/39.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	96	71.0/66.0	71.0/66.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	97	59.8/52.9	59.8/52.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	98	56.3/55.2	56.3/55.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	99	54.0/49.0	54.0/49.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	100	HDP1.0	HDP1.0	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	101	45.6/47.1	45.6/47.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	102	45.7/42.3	45.7/42.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	103	53.9/46.9	53.9/46.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	104	74.5/66.9	74.5/66.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	105	58.5/54.4	58.5/54.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	106	50.1/44.6	50.1/44.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	107	65.1/60.6	65.1/60.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	108	64.3/56.5	64.3/56.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	109	71.5/70.3	71.5/70.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	110	55.8/50.7	55.8/50.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	111	HDP1.5	HDP1.5	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	112	49.6/50.4	49.6/50.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	113	58.7/54.4	58.7/54.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	114	53.2/48.2	53.2/48.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	115	74.3/67.1	74.3/67.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	116	57.4/54.5	57.4/54.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	117	50.6/45.0	50.6/45.0	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	118	70.0/64.7	70.0/64.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	119	65.5/57.2	65.5/57.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	120	69.6/68.6	69.6/68.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	121	55.6/50.3	55.6/50.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	122	HDP2.0	HDP2.0	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	161	123	66.4/65.1	66.4/65.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	124	56.5/49.5	56.5/49.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	125	54.2/46.4	54.2/46.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	126	71.6/64.1	71.6/64.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	127	51.7/48.3	51.7/48.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	128	49.4/43.3	49.4/43.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	129	76.3/70.5	76.3/70.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	130	70.7/62.9	70.7/62.9	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	131	74.1/73.3	74.1/73.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	132	54.4/48.5	54.4/48.5	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	133	+	+	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	161	134	/−	/−	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	135	-0.8/-1.7	-0.8/-1.7	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	136	+3.8/+2.5	+3.8/+2.5	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	137	-11.4/-15.4	-11.4/-15.4	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	161	138	+1.7/+3.5	+1.7/+3.5	ADD
work_j4jyyr4gnrbelgfykpcumrgui4	161	139	+6.7/+2.4	+6.7/+2.4	ADD
work_j4jyyr4gnrbelgfykpcumrgui4	161	140	-3.1/-3.9	-3.1/-3.9	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	141	+5.5/+3.3	+5.5/+3.3	NFP
work_j4jyyr4gnrbelgfykpcumrgui4	161	142	-0.5/-1.9	-0.5/-1.9	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	143	+12.7/+13.5	+12.7/+13.5	FW
work_j4jyyr4gnrbelgfykpcumrgui4	161	144	-2.5/-3.7	-2.5/-3.7	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	145	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	146	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	147	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	161	148	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	149	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	150	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	151	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	152	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	153	Argument	Argument	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	154	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	155	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	156	MLE	MLE	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	157	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	158	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	161	159	four	four	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	160	hyper	hyper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	161	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	161	162	parameter	parameter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	163	settings	setting	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	161	164	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	165	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	166	HDP-	HDP-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	167	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	168	against	against	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	169	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	170	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	171	formalisms	formalism	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	161	172	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	161	173	participated	participate	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	161	174	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	175	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	176	PASCAL	PASCAL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	177	Challenge	Challenge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	178	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	179	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	161	180	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	161	181	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	182	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	183	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	161	184	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	185	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	186	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	161	187	BH	BH	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	188	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	189	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	190	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	161	191	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	192	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	161	193	2012a	2012a	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	194	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	195	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	161	196	BC	BC	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	197	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	198	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	199	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	161	200	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	161	201	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	161	202	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	161	203	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	161	204	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	161	205	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	206	addition	addition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	207	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	208	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	209	max	max	NN
work_j4jyyr4gnrbelgfykpcumrgui4	161	210	over	over	IN
work_j4jyyr4gnrbelgfykpcumrgui4	161	211	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	161	212	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	161	213	participants	participant	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	161	214	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	162	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	162	2	trained	train	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	162	3	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	162	4	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	162	5	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	162	6	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	162	7	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	162	8	punctuation	punctuation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	162	9	removed	remove	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	162	10	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	162	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	162	12	including	include	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	162	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	162	14	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	162	15	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	162	16	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	162	17	recommended	recommend	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	162	18	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	162	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	162	20	organizers	organizer	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	162	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	163	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	163	2	last	last	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	163	3	row	row	NN
work_j4jyyr4gnrbelgfykpcumrgui4	163	4	indicates	indicate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	163	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	163	6	difference	difference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	163	7	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	163	8	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	163	9	best	good	JJS
work_j4jyyr4gnrbelgfykpcumrgui4	163	10	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	163	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	163	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	163	13	competition	competition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	163	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	164	1	global	global	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	164	2	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	164	3	distribution	distribution	NN
work_j4jyyr4gnrbelgfykpcumrgui4	164	4	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	164	5	influence	influence	VB
work_j4jyyr4gnrbelgfykpcumrgui4	164	6	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	164	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	164	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	164	9	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	164	10	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	164	11	distributions	distribution	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	164	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	165	1	Further	further	RB
work_j4jyyr4gnrbelgfykpcumrgui4	165	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	165	3	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	165	4	provides	provide	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	165	5	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	165	6	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	165	7	simple	simple	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	165	8	knob	knob	NN
work_j4jyyr4gnrbelgfykpcumrgui4	165	9	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	165	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	165	11	choice	choice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	165	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	165	13	hyperparame-	hyperparame-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	165	14	ters	ter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	165	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	165	16	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	165	17	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	165	18	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	165	19	substantial	substantial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	165	20	effect	effect	NN
work_j4jyyr4gnrbelgfykpcumrgui4	165	21	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	165	22	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	165	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	166	1	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	166	2	side	side	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	166	3	effect	effect	NN
work_j4jyyr4gnrbelgfykpcumrgui4	166	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	166	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	166	6	hyperparameters	hyperparameter	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	166	7	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	166	8	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	166	9	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	166	10	strength	strength	NN
work_j4jyyr4gnrbelgfykpcumrgui4	166	11	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	166	12	determines	determine	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	166	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	166	14	rate	rate	NN
work_j4jyyr4gnrbelgfykpcumrgui4	166	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	166	16	convergence	convergence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	166	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	167	1	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	167	2	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	167	3	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	167	4	one	one	CD
work_j4jyyr4gnrbelgfykpcumrgui4	167	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	6	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	167	7	reasons	reason	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	167	8	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	167	10	high	high	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	167	11	vari-	vari-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	167	12	ance	ance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	167	13	seen	see	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	167	14	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	15	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	167	16	four	four	CD
work_j4jyyr4gnrbelgfykpcumrgui4	167	17	settings	setting	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	167	18	tested	test	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	167	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	167	20	although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	21	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	167	22	note	note	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	167	23	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	24	since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	25	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	167	26	initialization	initialization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	167	27	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	167	28	always	always	RB
work_j4jyyr4gnrbelgfykpcumrgui4	167	29	uniform	uniform	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	167	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	167	31	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	167	32	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	167	33	random	random	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	167	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	167	35	consecutive	consecutive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	167	36	runs	run	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	167	37	do	do	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	167	38	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	167	39	introduce	introduce	VB
work_j4jyyr4gnrbelgfykpcumrgui4	167	40	variance	variance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	167	41	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	167	42	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	167	43	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	167	44	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	167	45	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	167	46	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	168	1	7.2	7.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	168	2	Comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	168	3	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	4	systems	system	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	168	5	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	168	6	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	168	7	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	168	8	constraints	constraint	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	168	9	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	10	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	168	11	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	168	12	algorithm	algorithm	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	168	13	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	168	14	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	168	15	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	168	17	knowl-	knowl-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	168	18	edge	edge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	168	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	20	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	168	21	POS	POS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	168	22	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	168	23	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	168	24	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	168	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	168	26	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	168	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	168	28	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	168	29	compare	compare	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	168	30	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	31	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	168	32	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	168	33	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	168	34	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	168	35	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	168	36	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	168	37	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	168	38	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	168	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	169	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	169	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	169	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	169	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	169	5	who	who	WP
work_j4jyyr4gnrbelgfykpcumrgui4	169	6	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	169	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	169	8	nonparametric	nonparametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	169	9	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	169	10	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	169	11	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	169	12	incorporates	incorporate	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	169	13	thirteen	thirteen	CD
work_j4jyyr4gnrbelgfykpcumrgui4	169	14	universal	universal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	169	15	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	169	16	constraints	constraint	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	169	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	170	1	Three	three	CD
work_j4jyyr4gnrbelgfykpcumrgui4	170	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	170	3	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	170	4	constraints	constraint	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	5	correspond	correspond	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	170	6	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	170	7	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	170	8	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	9	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	170	10	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	11	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	170	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	170	13	roots	root	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	170	15	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	170	17	may	may	MD
work_j4jyyr4gnrbelgfykpcumrgui4	170	18	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	170	19	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	20	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	170	21	dependents	dependent	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	170	23	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	170	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	170	25	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	170	26	ten	ten	CD
work_j4jyyr4gnrbelgfykpcumrgui4	170	27	con-	con-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	170	28	straints	straint	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	170	29	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	170	30	e.g.	e.g.	RB
work_j4jyyr4gnrbelgfykpcumrgui4	171	1	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	171	2	adjectives	adjective	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	171	3	modify	modify	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	171	4	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	171	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	171	6	adverbs	adverbs	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	171	7	modify	modify	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	171	8	adjectives	adjective	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	171	9	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	171	10	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	171	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	171	12	etc	etc	FW
work_j4jyyr4gnrbelgfykpcumrgui4	171	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	171	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	172	1	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	172	2	no	no	DT
work_j4jyyr4gnrbelgfykpcumrgui4	172	3	equivalent	equivalent	NN
work_j4jyyr4gnrbelgfykpcumrgui4	172	4	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	172	5	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	172	6	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	172	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	173	1	Although	although	IN
work_j4jyyr4gnrbelgfykpcumrgui4	173	2	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	173	3	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	173	4	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	173	5	less	less	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	173	6	prior	prior	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	173	7	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	173	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	173	9	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	173	10	still	still	RB
work_j4jyyr4gnrbelgfykpcumrgui4	173	11	performs	perform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	173	12	competitively	competitively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	173	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	174	1	On	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	174	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	174	3	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	174	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	174	5	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	174	6	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	174	7	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	174	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	175	1	demonstrate	demonstrate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	175	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	175	3	im-	im-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	4	portance	portance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	5	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	175	6	effect	effect	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	175	9	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	10	choice	choice	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	12	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	13	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	175	14	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	15	comparing	compare	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	175	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	175	17	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	18	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	19	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	175	20	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	21	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	22	hand	hand	NN
work_j4jyyr4gnrbelgfykpcumrgui4	175	23	crafted	craft	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	175	24	universal	universal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	25	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	175	26	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	175	27	71.9	71.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	175	28	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	175	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	175	30	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	31	En-	en-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	32	glish	glish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	33	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	175	34	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	175	35	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	175	36	73.8	73.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	175	37	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	175	38	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	175	39	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	175	40	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	41	rules	rule	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	175	42	proposed	propose	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	175	43	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	175	44	Druck	Druck	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	175	45	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	175	46	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	175	47	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	176	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	176	2	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	176	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	176	4	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	176	5	64.9	64.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	176	6	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	176	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	177	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	177	2	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	177	4	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	5	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	6	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	7	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	8	’s	’s	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	9	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	10	drops	drop	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	177	11	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	177	12	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	177	13	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	177	14	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	15	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	16	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	177	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	177	18	presumable	presumable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	177	19	parse	parse	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	20	complexity	complexity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	21	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	177	22	82	82	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	23	Sl	sl	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	24	Es	Es	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	25	Da	Da	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	26	Pt	Pt	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	27	Sv	Sv	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	28	∼#Tokens	∼#token	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	177	29	3.8	3.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	30	K	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	31	4.2	4.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	32	K	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	33	9.5	9.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	34	K	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	35	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	36	K	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	37	24	24	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	38	K	k	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	39	N10	N10	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	40	50.9	50.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	41	67.2	67.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	42	51.9	51.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	43	71.5	71.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	44	63.3	63.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	45	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	46	56.6	56.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	47	62.1	62.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	48	51.5	51.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	49	74.7	74.7	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	50	69.8	69.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	51	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	52	2	2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	177	53	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	177	54	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	177	55	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	56	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	177	57	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	177	58	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	177	59	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	177	60	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	61	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	62	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	177	63	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	178	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	178	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	178	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	178	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	178	5	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	178	6	trained	trained	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	178	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	178	8	tested	test	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	178	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	178	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	178	11	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	178	12	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	178	13	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	178	14	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	178	15	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	178	16	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	178	17	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	178	18	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	178	19	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	178	20	Shared	Shared	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	178	21	Task	Task	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	178	22	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	179	1	increases	increase	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	179	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	179	3	whereas	whereas	IN
work_j4jyyr4gnrbelgfykpcumrgui4	179	4	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	179	5	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	179	6	shows	show	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	179	7	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	179	8	less	less	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	179	9	decline	decline	NN
work_j4jyyr4gnrbelgfykpcumrgui4	179	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	179	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	179	12	outperforms	outperform	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	179	13	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	179	14	universal	universal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	179	15	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	179	16	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	179	17	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	179	18	significant	significant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	179	19	margin.7	margin.7	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	20	≤	≤	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	21	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	22	≤	≤	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	179	23	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	24	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	25	Universal	Universal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	26	Rules	Rules	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	179	27	71.9	71.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	28	50.4	50.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	29	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	30	English	English	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	31	Rules	Rules	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	179	32	73.8	73.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	33	66.1	66.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	34	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	35	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	179	36	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	37	68.2	68.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	38	64.2	64.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	39	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	40	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	179	41	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	42	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	179	43	train	train	NN
work_j4jyyr4gnrbelgfykpcumrgui4	179	44	≤	≤	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	45	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	46	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	179	47	71.9	71.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	179	48	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	179	49	contrast	contrast	NN
work_j4jyyr4gnrbelgfykpcumrgui4	179	50	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	179	51	Spitkovsky	Spitkovsky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	52	et	et	FW
work_j4jyyr4gnrbelgfykpcumrgui4	179	53	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	179	54	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	180	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	180	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	180	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	180	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	180	5	who	who	WP
work_j4jyyr4gnrbelgfykpcumrgui4	180	6	reported	report	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	180	7	that	that	DT
work_j4jyyr4gnrbelgfykpcumrgui4	180	8	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	180	10	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	180	11	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	12	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	180	13	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	14	degrades	degrade	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	180	15	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	180	16	trained	train	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	180	17	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	180	18	longer	long	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	180	19	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	180	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	180	21	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	180	22	per-	per-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	23	formance	formance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	24	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	180	25	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	180	26	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	180	27	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	180	28	increases	increase	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	180	29	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	180	30	71.9	71.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	180	31	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	180	32	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	180	33	train	train	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	180	34	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	180	35	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	180	36	up	up	RP
work_j4jyyr4gnrbelgfykpcumrgui4	180	37	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	180	38	length	length	VB
work_j4jyyr4gnrbelgfykpcumrgui4	180	39	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	180	40	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	181	1	Another	another	DT
work_j4jyyr4gnrbelgfykpcumrgui4	181	2	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	181	3	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	181	4	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	181	5	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	181	6	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	181	7	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	8	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	181	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	181	10	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	181	11	captures	capture	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	181	12	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	181	13	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	181	14	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	181	15	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	181	16	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	17	ours	our	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	181	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	181	19	was	be	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	181	20	presented	present	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	181	21	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	22	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	181	23	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	181	24	Steed-	Steed-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	181	25	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	181	26	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	181	27	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	181	28	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	181	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	181	30	who	who	WP
work_j4jyyr4gnrbelgfykpcumrgui4	181	31	achieve	achieve	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	181	32	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	181	33	accuracy	accuracy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	181	34	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	35	74.5	74.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	181	36	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	37	WSJ10	WSJ10	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	181	38	section	section	NN
work_j4jyyr4gnrbelgfykpcumrgui4	181	39	23	23	CD
work_j4jyyr4gnrbelgfykpcumrgui4	181	40	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	181	41	trained	train	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	181	42	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	181	43	sections	section	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	181	44	02	02	CD
work_j4jyyr4gnrbelgfykpcumrgui4	181	45	-	-	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	181	46	22	22	CD
work_j4jyyr4gnrbelgfykpcumrgui4	181	47	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	181	48	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	182	1	Us-	us-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	182	2	ing	e	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	182	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	182	4	same	same	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	182	5	settings	setting	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	182	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	182	7	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	182	8	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	182	9	achieves	achieve	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	182	10	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	182	11	accu-	accu-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	182	12	racy	racy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	182	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	182	14	68.4	68.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	182	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	183	1	Unlike	unlike	IN
work_j4jyyr4gnrbelgfykpcumrgui4	183	2	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	183	3	approach	approach	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	183	5	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	183	6	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	183	7	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	183	8	do	do	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	183	9	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	183	10	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	183	11	induce	induce	VB
work_j4jyyr4gnrbelgfykpcumrgui4	183	12	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	183	13	appropri-	appropri-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	14	ate	ate	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	15	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	16	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	183	17	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	18	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	183	20	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	183	21	use	use	VB
work_j4jyyr4gnrbelgfykpcumrgui4	183	22	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	183	23	exten-	exten-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	24	sive	sive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	25	questionnaire	questionnaire	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	26	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	183	27	defines	define	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	183	28	prototype	prototype	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	183	29	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	183	30	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	183	31	various	various	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	32	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	33	constructions	construction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	183	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	183	35	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	183	36	requires	require	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	183	37	sig-	sig-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	38	nificant	nificant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	39	manual	manual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	40	engineering	engineering	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	41	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	183	42	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	183	43	POS	POS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	183	44	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	183	45	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	183	46	mapped	map	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	183	47	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	183	48	what	what	WP
work_j4jyyr4gnrbelgfykpcumrgui4	183	49	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	183	50	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	183	51	generate	generate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	183	52	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	183	53	language-	language-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	54	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	183	55	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	183	56	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	184	1	However	however	RB
work_j4jyyr4gnrbelgfykpcumrgui4	184	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	184	3	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	184	4	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	184	5	de-	de-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	184	6	grades	grade	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	184	7	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	184	8	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	184	9	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	184	10	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	184	11	subset	subset	NN
work_j4jyyr4gnrbelgfykpcumrgui4	184	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	184	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	184	14	ques-	ques-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	184	15	tions	tion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	184	16	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	184	17	considered	consider	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	184	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	185	1	Using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	185	2	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	185	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	185	4	first	first	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	5	14	14	CD
work_j4jyyr4gnrbelgfykpcumrgui4	185	6	ques-	ques-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	185	7	tions	tion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	9	covering	cover	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	185	10	facts	fact	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	11	about	about	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	185	13	ordering	ordering	NN
work_j4jyyr4gnrbelgfykpcumrgui4	185	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	15	subjects	subject	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	17	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	185	19	objects	object	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	21	adjectives	adjective	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	23	adverbs	adverb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	25	auxiliaries	auxiliary	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	27	adpositions	adposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	28	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	29	possessives	possessive	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	30	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	185	31	relative	relative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	32	markers	marker	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	33	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	34	they	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	185	35	achieve	achieve	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	185	36	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	185	37	accuracy	accuracy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	185	38	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	39	68.2	68.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	185	40	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	41	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	185	42	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	185	43	almost	almost	RB
work_j4jyyr4gnrbelgfykpcumrgui4	185	44	iden-	iden-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	45	7Our	7our	CD
work_j4jyyr4gnrbelgfykpcumrgui4	185	46	earlier	early	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	185	47	generative	generative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	48	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	185	49	showed	show	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	185	50	similar	similar	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	51	behavior	behavior	NN
work_j4jyyr4gnrbelgfykpcumrgui4	185	52	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	185	53	al-	al-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	185	54	though	though	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	55	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	185	56	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	57	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	58	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	185	59	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	185	60	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	185	61	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	185	62	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	185	63	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	185	64	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	185	65	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	185	66	directly	directly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	185	67	comparable	comparable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	185	68	due	due	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	69	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	70	differences	difference	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	71	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	185	72	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	185	73	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	185	74	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	186	1	Sl	Sl	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	2	Es	Es	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	3	Da	Da	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	4	Pt	Pt	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	5	Sv	Sv	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	6	#	#	$
work_j4jyyr4gnrbelgfykpcumrgui4	186	7	Tokens	token	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	186	8	3,857	3,857	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	9	4,230	4,230	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	10	9,549	9,549	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	11	15,015	15,015	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	12	24,021	24,021	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	13	G10	G10	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	14	51.2	51.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	15	62.4	62.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	16	47.2	47.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	17	54.3	54.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	18	48.6	48.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	19	HDP	HDP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	20	57.9	57.9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	21	65.4	65.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	22	49.3	49.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	23	73.5	73.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	24	73.2	73.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	25	Bg	Bg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	26	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	27	Nl	Nl	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	28	Ja	Ja	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	29	De	De	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	30	#	#	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	31	Tokens	Tokens	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	32	38,220	38,220	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	33	42,442	42,442	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	34	43,405	43,405	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	35	43,501	43,501	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	36	77,705	77,705	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	37	G10	G10	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	38	59.8	59.8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	39	64.4	64.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	40	47.5	47.5	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	41	60.2	60.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	42	47.4	47.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	43	HDP	hdp	NN
work_j4jyyr4gnrbelgfykpcumrgui4	186	44	66.1	66.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	45	70.3	70.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	46	56.2	56.2	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	47	64.1	64.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	48	68.4	68.4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	49	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	186	50	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	186	51	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	186	52	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	186	53	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	186	54	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	186	55	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	186	56	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	186	57	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	186	58	Gillenwa-	Gillenwa-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	59	ter	ter	NN
work_j4jyyr4gnrbelgfykpcumrgui4	186	60	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	61	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	186	62	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	187	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	187	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	187	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	187	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	187	5	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	187	6	trained	train	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	187	7	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	187	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	187	9	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	10	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	187	11	train-	train-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	12	ing	ing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	13	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	187	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	187	15	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	187	16	tested	test	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	187	17	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	187	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	187	19	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	20	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	187	21	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	22	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	187	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	187	24	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	187	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	187	26	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	187	27	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	187	28	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	187	29	Shared	shared	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	187	30	task	task	NN
work_j4jyyr4gnrbelgfykpcumrgui4	187	31	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	188	1	tical	tical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	188	2	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	188	3	ours	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	188	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	188	5	even	even	RB
work_j4jyyr4gnrbelgfykpcumrgui4	188	6	though	though	IN
work_j4jyyr4gnrbelgfykpcumrgui4	188	7	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	188	8	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	188	9	significantly	significantly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	188	10	less	less	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	188	11	initial	initial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	188	12	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	188	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	189	1	However	however	RB
work_j4jyyr4gnrbelgfykpcumrgui4	189	2	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	189	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	189	4	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	189	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	189	6	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	189	7	below	below	RB
work_j4jyyr4gnrbelgfykpcumrgui4	189	8	indicate	indicate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	189	9	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	10	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	189	11	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	189	12	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	13	fact	fact	NN
work_j4jyyr4gnrbelgfykpcumrgui4	189	14	learning	learn	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	189	15	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	189	16	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	189	18	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	189	19	exact	exact	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	189	20	details	detail	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	189	21	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	22	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	23	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	189	24	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	189	25	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	189	26	con-	con-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	189	27	structed	structe	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	189	28	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	189	29	hand	hand	NN
work_j4jyyr4gnrbelgfykpcumrgui4	189	30	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	190	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	190	2	remaining	remain	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	190	3	14	14	CD
work_j4jyyr4gnrbelgfykpcumrgui4	190	4	questions	question	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	190	5	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	190	6	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	190	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	190	8	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	190	9	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	190	10	questionnaire	questionnaire	NN
work_j4jyyr4gnrbelgfykpcumrgui4	190	11	cover	cover	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	190	12	less	less	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	190	13	frequent	frequent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	14	phenomena	phenomenon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	190	15	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	16	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	190	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	190	18	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	190	19	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	190	20	negative	negative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	21	markers	marker	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	190	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	190	23	dative	dative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	24	shift	shift	NN
work_j4jyyr4gnrbelgfykpcumrgui4	190	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	190	26	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	190	27	pro	pro	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	28	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	190	29	drop	drop	NN
work_j4jyyr4gnrbelgfykpcumrgui4	190	30	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	191	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	191	2	obvious	obvious	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	191	3	ad-	ad-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	191	4	vantage	vantage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	191	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	191	6	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	191	7	approach	approach	NN
work_j4jyyr4gnrbelgfykpcumrgui4	191	8	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	191	9	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	191	10	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	191	11	allows	allow	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	191	12	them	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	191	13	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	191	14	define	define	VB
work_j4jyyr4gnrbelgfykpcumrgui4	191	15	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	191	16	much	much	RB
work_j4jyyr4gnrbelgfykpcumrgui4	191	17	more	more	RBR
work_j4jyyr4gnrbelgfykpcumrgui4	191	18	fine	fine	RB
work_j4jyyr4gnrbelgfykpcumrgui4	191	19	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	191	20	grained	grain	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	191	21	inventory	inventory	NN
work_j4jyyr4gnrbelgfykpcumrgui4	191	22	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	191	23	lexical	lexical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	191	24	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	191	25	than	than	IN
work_j4jyyr4gnrbelgfykpcumrgui4	191	26	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	191	27	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	191	28	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	191	29	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	191	30	induce	induce	VB
work_j4jyyr4gnrbelgfykpcumrgui4	191	31	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	192	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	192	2	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	192	3	stipulate	stipulate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	192	4	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	5	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	6	certain	certain	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	7	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	192	8	knowl-	knowl-	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	192	9	edge	edge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	10	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	11	pro	pro	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	12	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	13	drop	drop	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	14	could	could	MD
work_j4jyyr4gnrbelgfykpcumrgui4	192	15	play	play	VB
work_j4jyyr4gnrbelgfykpcumrgui4	192	16	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	17	significant	significant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	18	role	role	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	19	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	21	success	success	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	22	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	23	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	192	24	approach	approach	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	25	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	192	26	if	if	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	27	complete	complete	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	28	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	192	29	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	192	30	allowed	allow	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	192	31	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	192	32	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	192	33	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	34	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	35	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	36	S\N	s\n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	37	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	192	38	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	192	39	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	192	40	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	192	41	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	192	42	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	43	same	same	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	44	lex-	lex-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	45	ical	ical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	46	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	47	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	192	48	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	192	49	used	use	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	192	50	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	51	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	52	verb	verb	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	53	regardless	regardless	RB
work_j4jyyr4gnrbelgfykpcumrgui4	192	54	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	55	whether	whether	IN
work_j4jyyr4gnrbelgfykpcumrgui4	192	56	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	192	57	subject	subject	NN
work_j4jyyr4gnrbelgfykpcumrgui4	192	58	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	192	59	present	present	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	192	60	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	192	61	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	192	62	been	be	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	192	63	dropped	drop	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	192	64	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	193	1	7.3	7.3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	193	2	Additional	additional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	193	3	Languages	Languages	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	193	4	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	193	5	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	193	6	to	to	TO
work_j4jyyr4gnrbelgfykpcumrgui4	193	7	provide	provide	VB
work_j4jyyr4gnrbelgfykpcumrgui4	193	8	results	result	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	193	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	193	10	additional	additional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	193	11	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	193	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	193	13	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	193	14	present	present	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	193	15	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	193	16	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	193	17	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	193	18	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	193	19	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	193	20	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	193	21	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	193	22	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	193	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	193	24	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	193	25	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	193	26	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	193	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	194	1	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	194	3	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	4	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	5	G10	G10	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	6	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	194	8	using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	194	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	194	10	ConLL	ConLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	11	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	194	12	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	13	Shared	share	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	194	14	Task	Task	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	15	data	datum	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	194	16	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	17	Buchholz	Buchholz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	194	19	Marsi	Marsi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	194	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	194	21	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	194	22	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	194	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	195	1	Fol-	Fol-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	195	2	lowing	low	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	195	3	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	195	4	et	et	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	195	5	al	al	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	195	6	.	.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	195	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	195	8	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	195	9	train	train	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	195	10	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	195	11	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	195	12	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	195	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	195	14	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	195	15	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	195	16	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	195	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	195	18	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	195	19	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	195	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	195	21	evaluate	evaluate	VB
work_j4jyyr4gnrbelgfykpcumrgui4	195	22	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	195	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	195	24	test	test	NN
work_j4jyyr4gnrbelgfykpcumrgui4	195	25	set	set	NN
work_j4jyyr4gnrbelgfykpcumrgui4	195	26	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	196	1	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	3	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	196	4	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	5	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	196	6	training	training	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	7	regime	regime	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	196	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	196	10	these	these	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	11	corpora	corpora	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	12	differ	differ	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	196	13	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	14	many	many	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	196	15	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	196	16	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	17	that	that	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	18	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	20	PASCAL	PASCAL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	196	21	challenge	challenge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	196	23	numbers	number	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	196	24	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	25	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	26	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	196	27	can-	can-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	28	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	196	29	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	196	30	compared	compare	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	196	31	directly	directly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	196	32	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	33	those	those	DT
work_j4jyyr4gnrbelgfykpcumrgui4	196	34	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	196	35	Table	table	NN
work_j4jyyr4gnrbelgfykpcumrgui4	196	36	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	196	37	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	197	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	197	2	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	197	3	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	197	4	applied	apply	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	197	5	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	197	6	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	197	7	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	197	8	Goldberg	Goldberg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	197	9	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	10	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	11	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	12	’s	’s	POS
work_j4jyyr4gnrbelgfykpcumrgui4	197	13	Hebrew	hebrew	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	197	14	corpus	corpus	NN
work_j4jyyr4gnrbelgfykpcumrgui4	197	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	197	16	where	where	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	197	17	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	197	18	achieves	achieve	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	197	19	an	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	197	20	accuracy	accuracy	NN
work_j4jyyr4gnrbelgfykpcumrgui4	197	21	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	197	22	62.1	62.1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	23	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	24	trained	train	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	197	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	197	26	tested	test	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	197	27	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	197	28	all	all	DT
work_j4jyyr4gnrbelgfykpcumrgui4	197	29	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	197	30	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	197	31	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	32	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	197	33	7,253	7,253	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	34	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	35	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	197	36	59.6	59.6	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	37	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	38	length	length	NN
work_j4jyyr4gnrbelgfykpcumrgui4	197	39	15	15	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	40	;	;	:
work_j4jyyr4gnrbelgfykpcumrgui4	197	41	21,422	21,422	CD
work_j4jyyr4gnrbelgfykpcumrgui4	197	42	tokens	token	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	197	43	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	197	44	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	198	1	83	83	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	2	Arabic	arabic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	3	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	4	Swedish	swedish	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	5	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	6	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	7	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	8	Childes	Childes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	9	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	10	Japanese	japanese	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	11	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	12	Czech	czech	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	13	%	%	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	14	VERB	VERB	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	15	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	16	S\N)/N	S\N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	17	56	56	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	18	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	19	45	45	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	20	S\N	s\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	21	52	52	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	22	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	23	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	24	N	N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	25	44	44	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	26	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	27	84	84	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	28	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	29	26	26	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	30	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	31	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	32	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	33	N)/N	N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	34	29	29	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	35	S\N	s\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	36	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	37	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	38	S\N)/N	S\N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	39	19	19	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	40	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	41	37	37	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	42	S\N	s\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	43	25	25	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	44	ADP	ADP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	45	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	46	68	68	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	47	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	48	S\S)/N	S\S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	49	49	49	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	50	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	51	S\S)/N	S\S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	52	46	46	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	53	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	54	S\S)/N	S\S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	55	45	45	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	56	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	57	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	58	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	59	S)\N	S)\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	60	44	44	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	61	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	62	S\S)/N	S\S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	63	42	42	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	64	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	65	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	66	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	67	21	21	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	68	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	69	N\N)/N	N\N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	70	25	25	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	71	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	72	N\N)/N	N\N)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	73	20	20	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	74	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	75	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	76	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	77	25	25	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	78	N\N	n\n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	79	23	23	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	80	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	198	81	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	82	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	83	S)/N	S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	84	26	26	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	85	NOUN	NOUN	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	86	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	87	50	50	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	88	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	89	91	91	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	90	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	91	79	79	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	92	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	93	89	89	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	94	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	95	73	73	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	96	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	97	74	74	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	98	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	99	35	35	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	100	N	n	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	101	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	102	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	103	12	12	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	104	ADJ	adj	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	105	N\N	N\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	106	82	82	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	107	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	108	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	109	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	110	50	50	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	111	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	112	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	113	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	114	70	70	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	115	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	116	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	117	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	118	46	46	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	119	S	s	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	120	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	121	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	198	122	64	64	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	123	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	124	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	198	125	N	n	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	126	55	55	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	127	Figure	figure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	128	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	129	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	198	130	Partial	partial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	131	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	198	132	demonstrating	demonstrate	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	198	133	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	134	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	198	135	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	136	learned	learn	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	198	137	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	198	138	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	198	139	five	five	CD
work_j4jyyr4gnrbelgfykpcumrgui4	198	140	lan-	lan-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	198	141	guages	guage	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	198	142	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	199	1	For	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	199	2	ease	ease	NN
work_j4jyyr4gnrbelgfykpcumrgui4	199	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	199	4	comparison	comparison	NN
work_j4jyyr4gnrbelgfykpcumrgui4	199	5	between	between	IN
work_j4jyyr4gnrbelgfykpcumrgui4	199	6	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	199	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	199	8	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	199	9	use	use	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	199	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	199	11	universal	universal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	199	12	tag	tag	NN
work_j4jyyr4gnrbelgfykpcumrgui4	199	13	label	label	NN
work_j4jyyr4gnrbelgfykpcumrgui4	199	14	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	199	15	Verb	Verb	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	199	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	199	17	Adposition	Adposition	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	199	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	199	19	Noun	Noun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	199	20	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	199	21	Adjective	Adjective	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	199	22	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	199	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	200	1	Shown	show	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	200	2	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	200	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	200	4	most	most	RBS
work_j4jyyr4gnrbelgfykpcumrgui4	200	5	common	common	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	200	6	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	200	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	200	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	200	9	fraction	fraction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	200	10	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	200	11	occurrences	occurrence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	200	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	200	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	200	14	tag	tag	NN
work_j4jyyr4gnrbelgfykpcumrgui4	200	15	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	200	16	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	200	17	assigned	assign	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	200	18	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	200	19	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	200	20	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	200	21	according	accord	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	200	22	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	200	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	200	24	Viterbi	Viterbi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	200	25	parses	parse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	200	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	200	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	201	1	8	8	CD
work_j4jyyr4gnrbelgfykpcumrgui4	201	2	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	201	3	Induced	induced	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	201	4	Lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	201	5	Since	since	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	6	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	201	7	approach	approach	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	8	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	201	9	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	201	10	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	11	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	201	12	lexicalized	lexicalized	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	201	13	for-	for-	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	14	malism	malism	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	15	such	such	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	201	16	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	17	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	201	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	201	19	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	201	20	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	21	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	201	22	in-	in-	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	201	23	duces	duce	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	201	24	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	201	25	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	201	26	pair	pair	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	201	27	words	word	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	201	28	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	201	29	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	201	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	201	31	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	32	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	201	33	case	case	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	201	35	POS-	POS-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	201	36	tags	tag	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	201	37	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	201	38	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	201	39	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	40	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	201	41	specific	specific	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	201	42	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	201	43	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	201	44	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	201	45	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	201	46	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	201	47	behavior	behavior	NN
work_j4jyyr4gnrbelgfykpcumrgui4	201	48	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	202	1	If	if	IN
work_j4jyyr4gnrbelgfykpcumrgui4	202	2	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	202	3	approach	approach	NN
work_j4jyyr4gnrbelgfykpcumrgui4	202	4	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	202	5	success-	success-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	202	6	ful	ful	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	202	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	202	8	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	202	9	should	should	MD
work_j4jyyr4gnrbelgfykpcumrgui4	202	10	learn	learn	VB
work_j4jyyr4gnrbelgfykpcumrgui4	202	11	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	202	12	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	202	13	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	202	14	properties	property	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	202	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	202	16	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	202	17	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	202	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	202	19	which	which	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	202	20	will	will	MD
work_j4jyyr4gnrbelgfykpcumrgui4	202	21	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	202	22	reflected	reflect	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	202	23	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	202	24	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	202	25	corre-	corre-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	202	26	sponding	sponde	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	202	27	lexicon	lexicon	NN
work_j4jyyr4gnrbelgfykpcumrgui4	202	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	203	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	203	2	Figure	figure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	3	3	3	CD
work_j4jyyr4gnrbelgfykpcumrgui4	203	4	one	one	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	5	sees	see	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	203	6	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	203	7	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	203	8	subcategorize	subcategorize	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	203	9	differently	differently	RB
work_j4jyyr4gnrbelgfykpcumrgui4	203	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	203	11	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	203	12	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	13	ordering	ordering	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	14	differs	differ	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	203	15	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	203	16	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	203	18	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	203	19	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	203	20	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	203	21	attachment	attachment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	203	22	structures	structure	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	203	23	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	203	24	prepositions	preposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	203	25	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	203	26	automatically	automatically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	203	27	discovered	discover	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	203	28	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	203	29	differ	differ	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	203	30	across	across	IN
work_j4jyyr4gnrbelgfykpcumrgui4	203	31	languages	language	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	203	32	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	204	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	204	2	Arabic	Arabic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	204	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	204	4	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	204	5	example	example	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	204	7	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	204	8	sys-	sys-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	204	9	tem	tem	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	10	learns	learn	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	204	11	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	204	12	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	13	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	14	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	204	15	variable	variable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	204	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	204	17	therefore	therefore	RB
work_j4jyyr4gnrbelgfykpcumrgui4	204	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	204	19	verb	verb	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	20	must	must	MD
work_j4jyyr4gnrbelgfykpcumrgui4	204	21	allow	allow	VB
work_j4jyyr4gnrbelgfykpcumrgui4	204	22	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	204	23	both	both	CC
work_j4jyyr4gnrbelgfykpcumrgui4	204	24	SVO	SVO	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	204	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	204	26	VOS	VOS	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	204	27	style	style	NN
work_j4jyyr4gnrbelgfykpcumrgui4	204	28	constructions	construction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	204	29	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	205	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	205	2	generally	generally	RB
work_j4jyyr4gnrbelgfykpcumrgui4	205	3	learn	learn	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	205	4	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	205	5	adpositions	adposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	205	6	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	205	7	prepositions	preposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	205	8	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	205	9	postpositions	postposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	205	10	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	205	11	take	take	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	205	12	nouns	noun	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	205	13	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	205	14	argu-	argu-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	205	15	ments	ment	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	205	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	206	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	206	2	Czech	Czech	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	206	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	206	4	PPs	pp	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	206	5	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	206	6	appear	appear	VB
work_j4jyyr4gnrbelgfykpcumrgui4	206	7	before	before	RB
work_j4jyyr4gnrbelgfykpcumrgui4	206	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	206	9	after	after	IN
work_j4jyyr4gnrbelgfykpcumrgui4	206	10	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	206	11	verb	verb	NN
work_j4jyyr4gnrbelgfykpcumrgui4	206	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	206	13	leading	lead	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	206	14	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	206	15	two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	206	16	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	206	17	categories	category	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	206	18	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	206	19	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	206	20	S\S)/N	S\S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	206	21	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	206	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	206	23	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	206	24	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	206	25	S)/N	S)/N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	206	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	206	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	207	1	Japanese	Japanese	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	2	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	207	3	postpositions	postposition	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	4	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	207	5	ap-	ap-	CD
work_j4jyyr4gnrbelgfykpcumrgui4	207	6	pear	pear	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	7	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	8	preverbal	preverbal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	207	9	position	position	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	10	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	11	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	12	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	13	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	207	14	S)\N	S)\N	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	15	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	207	17	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	207	18	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	207	19	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	20	category	category	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	21	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	207	22	assigned	assign	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	207	23	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	24	nominal	nominal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	207	25	particles	particle	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	26	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	207	27	cor-	cor-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	207	28	respond	respond	VB
work_j4jyyr4gnrbelgfykpcumrgui4	207	29	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	30	case	case	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	31	markers	marker	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	207	33	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	207	34	effectively	effectively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	207	35	absorbs	absorb	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	207	36	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	37	noun	noun	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	38	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	207	39	leading	lead	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	207	40	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	41	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	42	preference	preference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	43	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	44	verbs	verb	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	45	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	207	46	do	do	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	207	47	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	207	48	take	take	VB
work_j4jyyr4gnrbelgfykpcumrgui4	207	49	any	any	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	50	arguments	argument	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	51	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	52	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	53	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	54	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	207	55	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	207	56	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	57	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	58	misanalysis	misanalysis	NN
work_j4jyyr4gnrbelgfykpcumrgui4	207	59	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	60	ad-	ad-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	207	61	jectives	jective	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	62	as	as	IN
work_j4jyyr4gnrbelgfykpcumrgui4	207	63	verb	verb	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	207	64	modifiers	modifier	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	207	65	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	66	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	67	/	/	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	207	68	S	S	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	207	69	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	207	70	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	208	1	Our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	208	2	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	208	3	also	also	RB
work_j4jyyr4gnrbelgfykpcumrgui4	208	4	reflect	reflect	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	208	5	differences	difference	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	208	6	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	208	7	style	style	NN
work_j4jyyr4gnrbelgfykpcumrgui4	208	8	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	208	9	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	208	10	Childes	Childes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	208	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	208	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	208	13	WSJ	WSJ	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	208	14	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	208	15	both	both	DT
work_j4jyyr4gnrbelgfykpcumrgui4	208	16	English	English	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	208	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	208	18	they	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	208	19	represent	represent	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	208	20	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	208	21	different	different	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	208	22	registers	register	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	208	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	209	1	We	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	209	2	learn	learn	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	209	3	that	that	IN
work_j4jyyr4gnrbelgfykpcumrgui4	209	4	subjects	subject	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	209	5	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	209	6	mostly	mostly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	209	7	absent	absent	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	209	8	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	209	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	209	10	informal	informal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	209	11	speech	speech	NN
work_j4jyyr4gnrbelgfykpcumrgui4	209	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	209	13	child	child	NN
work_j4jyyr4gnrbelgfykpcumrgui4	209	14	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	209	15	directed	direct	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	209	16	instructions	instruction	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	209	17	contained	contain	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	209	18	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	209	19	Childes	Childes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	209	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	209	21	while	while	IN
work_j4jyyr4gnrbelgfykpcumrgui4	209	22	effectively	effectively	RB
work_j4jyyr4gnrbelgfykpcumrgui4	209	23	mandatory	mandatory	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	209	24	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	209	25	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	209	26	Wall	Wall	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	209	27	Street	Street	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	209	28	Journal	Journal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	209	29	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	210	1	9	9	CD
work_j4jyyr4gnrbelgfykpcumrgui4	210	2	Conclusions	conclusion	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	210	3	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	210	4	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	5	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	210	6	introduced	introduce	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	210	7	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	210	8	novel	novel	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	9	factorization	factorization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	10	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	210	11	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	210	12	models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	210	13	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	210	14	showed	show	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	210	15	how	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	210	16	when	when	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	210	17	combined	combine	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	210	18	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	210	19	non	non	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	20	-	-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	21	parametric	parametric	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	22	Bayesian	bayesian	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	23	statistics	statistic	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	210	24	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	210	25	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	210	26	compete	compete	VB
work_j4jyyr4gnrbelgfykpcumrgui4	210	27	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	210	28	every	every	DT
work_j4jyyr4gnrbelgfykpcumrgui4	210	29	other	other	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	30	grammar	grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	210	31	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	32	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	33	cur-	cur-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	34	rently	rently	RB
work_j4jyyr4gnrbelgfykpcumrgui4	210	35	available	available	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	36	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	210	37	including	include	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	210	38	those	those	DT
work_j4jyyr4gnrbelgfykpcumrgui4	210	39	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	210	40	capture	capture	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	210	41	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	210	42	sig-	sig-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	210	43	nificant	nificant	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	44	amount	amount	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	45	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	210	46	prior	prior	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	47	linguistic	linguistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	210	48	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	210	49	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	211	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	211	2	use	use	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	211	4	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	211	5	powerful	powerful	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	211	6	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	211	7	formalism	formalism	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	8	proves	prove	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	211	9	ben-	ben-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	211	10	eficial	eficial	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	211	11	both	both	CC
work_j4jyyr4gnrbelgfykpcumrgui4	211	12	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	211	13	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	211	14	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	211	15	requiring	require	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	211	16	very	very	RB
work_j4jyyr4gnrbelgfykpcumrgui4	211	17	limited	limited	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	211	18	uni-	uni-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	211	19	versal	versal	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	20	knowledge	knowledge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	21	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	211	22	robustness	robustness	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	23	at	at	IN
work_j4jyyr4gnrbelgfykpcumrgui4	211	24	longer	long	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	211	25	sentence	sentence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	211	26	lengths	length	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	211	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	212	1	Unlike	unlike	IN
work_j4jyyr4gnrbelgfykpcumrgui4	212	2	standard	standard	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	212	3	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	4	induction	induction	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	5	sys-	sys-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	212	6	tems	tem	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	212	7	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	212	8	are	be	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	212	9	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	212	10	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	212	11	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	12	grammar	grammar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	212	14	our	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	212	15	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	16	returns	return	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	212	17	linguistically	linguistically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	212	18	interpretable	interpretable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	212	19	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	212	20	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	212	21	each	each	DT
work_j4jyyr4gnrbelgfykpcumrgui4	212	22	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	23	that	that	WDT
work_j4jyyr4gnrbelgfykpcumrgui4	212	24	demonstrate	demonstrate	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	212	25	it	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	212	26	has	have	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	212	27	discov-	discov-	RB
work_j4jyyr4gnrbelgfykpcumrgui4	212	28	ered	ere	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	212	29	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	212	30	basic	basic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	212	31	word	word	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	32	order	order	NN
work_j4jyyr4gnrbelgfykpcumrgui4	212	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	213	1	Of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	213	2	particular	particular	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	213	3	note	note	NN
work_j4jyyr4gnrbelgfykpcumrgui4	213	4	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	213	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	213	6	simplicity	simplicity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	213	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	213	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	213	9	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	213	10	both	both	CC
work_j4jyyr4gnrbelgfykpcumrgui4	213	11	algorithmically	algorithmically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	213	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	213	13	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	213	14	terms	term	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	213	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	213	16	implementation	implementation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	213	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	214	1	By	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	214	2	not	not	RB
work_j4jyyr4gnrbelgfykpcumrgui4	214	3	faltering	falter	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	214	4	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	214	5	longer	long	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	214	6	sentences	sentence	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	214	7	or	or	CC
work_j4jyyr4gnrbelgfykpcumrgui4	214	8	requiring	require	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	214	9	extensive	extensive	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	214	10	tuning	tuning	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	214	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	214	13	system	system	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	14	can	can	MD
work_j4jyyr4gnrbelgfykpcumrgui4	214	15	be	be	VB
work_j4jyyr4gnrbelgfykpcumrgui4	214	16	easily	easily	RB
work_j4jyyr4gnrbelgfykpcumrgui4	214	17	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	214	18	quickly	quickly	RB
work_j4jyyr4gnrbelgfykpcumrgui4	214	19	deployed	deploy	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	214	20	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	214	21	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	214	22	new	new	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	214	23	lan-	lan-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	24	guage	guage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	25	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	214	26	return	return	VB
work_j4jyyr4gnrbelgfykpcumrgui4	214	27	state	state	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	28	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	214	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	214	30	art	art	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	31	performance	performance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	214	32	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	214	33	easily	easily	RB
work_j4jyyr4gnrbelgfykpcumrgui4	214	34	interpretable	interpretable	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	214	35	lexicons	lexicon	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	214	36	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	215	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	215	2	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	3	paper	paper	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	215	5	we	-PRON-	PRP
work_j4jyyr4gnrbelgfykpcumrgui4	215	6	have	have	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	215	7	applied	apply	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	215	8	this	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	9	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	10	only	only	RB
work_j4jyyr4gnrbelgfykpcumrgui4	215	11	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	215	12	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	13	restricted	restrict	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	215	14	fragment	fragment	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	15	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	215	16	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	215	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	215	18	but	but	CC
work_j4jyyr4gnrbelgfykpcumrgui4	215	19	future	future	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	215	20	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	21	will	will	MD
work_j4jyyr4gnrbelgfykpcumrgui4	215	22	address	address	VB
work_j4jyyr4gnrbelgfykpcumrgui4	215	23	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	24	impact	impact	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	25	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	215	26	lex-	lex-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	27	icalization	icalization	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	28	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	215	29	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	215	30	inclusion	inclusion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	215	31	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	215	32	richer	rich	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	215	33	combinators	combinator	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	215	34	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	216	1	10	10	CD
work_j4jyyr4gnrbelgfykpcumrgui4	216	2	Acknowledgements	acknowledgement	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	216	3	This	this	DT
work_j4jyyr4gnrbelgfykpcumrgui4	216	4	work	work	NN
work_j4jyyr4gnrbelgfykpcumrgui4	216	5	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	216	6	supported	support	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	216	7	by	by	IN
work_j4jyyr4gnrbelgfykpcumrgui4	216	8	NSF	NSF	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	9	CAREER	CAREER	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	10	award	award	NN
work_j4jyyr4gnrbelgfykpcumrgui4	216	11	1053856	1053856	CD
work_j4jyyr4gnrbelgfykpcumrgui4	216	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	216	13	Bayesian	Bayesian	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	14	Models	Models	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	15	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	216	16	Lexicalized	Lexicalized	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	17	Gram-	Gram-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	18	mars	mars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	216	19	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	216	20	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	217	1	84	84	CD
work_j4jyyr4gnrbelgfykpcumrgui4	217	2	References	References	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	217	3	Taylor	Taylor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	217	4	Berg	Berg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	217	5	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	217	6	Kirkpatrick	Kirkpatrick	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	217	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	217	8	Dan	Dan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	217	9	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	217	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	218	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	218	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	219	1	Phyloge-	Phyloge-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	219	2	netic	netic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	219	3	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	219	4	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	219	5	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	220	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	220	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	220	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	220	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	220	5	48th	48th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	220	6	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	7	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	220	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	220	10	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	11	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	220	12	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	13	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	220	15	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	220	16	1288–1297	1288–1297	CD
work_j4jyyr4gnrbelgfykpcumrgui4	220	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	220	18	Uppsala	Uppsala	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	220	20	Sweden	Sweden	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	220	22	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	220	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	221	1	Christopher	Christopher	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	221	2	Bishop	Bishop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	221	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	222	1	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	222	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	223	1	Pattern	Pattern	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	223	2	Recognition	Recognition	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	223	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	223	4	Ma-	Ma-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	223	5	chine	chine	NN
work_j4jyyr4gnrbelgfykpcumrgui4	223	6	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	223	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	224	1	Springer	Springer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	224	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	224	3	Verlag	Verlag	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	224	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	224	5	August	August	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	224	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	225	1	Yonatan	Yonatan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	225	2	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	225	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	225	4	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	225	5	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	225	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	226	1	2012a	2012a	CD
work_j4jyyr4gnrbelgfykpcumrgui4	226	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	227	1	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	227	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	227	3	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	227	4	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	227	5	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	227	6	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	227	7	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	227	8	Grammars	Grammars	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	227	9	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	228	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	228	2	NAACL	NAACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	3	HLT	HLT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	4	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	5	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	228	6	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	228	8	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	9	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	228	11	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	228	12	90–95	90–95	CD
work_j4jyyr4gnrbelgfykpcumrgui4	228	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	228	14	Montréal	Montréal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	228	16	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	228	18	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	228	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	229	1	Yonatan	Yonatan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	229	2	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	229	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	229	4	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	229	5	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	229	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	230	1	2012b	2012b	CD
work_j4jyyr4gnrbelgfykpcumrgui4	230	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	231	1	Simple	Simple	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	2	Robust	Robust	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	3	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	4	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	5	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	231	6	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	7	Cate-	Cate-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	8	gorial	gorial	NN
work_j4jyyr4gnrbelgfykpcumrgui4	231	9	Grammars	Grammars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	231	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	232	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	232	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	232	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	232	5	Twenty	Twenty	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	6	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	232	7	Sixth	Sixth	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	8	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	232	10	Artificial	Artificial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	11	Intelligence	Intelligence	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	232	13	AAAI-12	AAAI-12	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	232	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	232	16	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	232	17	1643–1649	1643–1649	CD
work_j4jyyr4gnrbelgfykpcumrgui4	232	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	232	19	Toronto	Toronto	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	232	21	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	232	23	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	232	24	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	233	1	David	David	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	2	M	M	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	3	Blei	Blei	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	233	5	Michael	Michael	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	6	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	7	Jordan	Jordan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	233	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	234	1	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	234	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	235	1	Variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	235	2	Methods	method	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	235	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	235	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	235	5	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	235	6	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	235	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	236	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	236	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	236	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	236	5	Twenty	Twenty	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	6	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	236	7	First	First	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	8	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	9	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	10	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	236	11	Machine	Machine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	12	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	13	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	236	14	ICML	ICML	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	15	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	236	16	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	236	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	236	18	Banff	Banff	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	236	20	Alberta	Alberta	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	236	22	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	236	24	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	236	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	237	1	Phil	Phil	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	237	2	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	237	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	237	4	Trevor	Trevor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	237	5	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	237	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	238	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	238	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	239	1	Unsupervised	unsupervise	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	239	2	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	239	4	Tree	Tree	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	5	Substitution	Substitution	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	6	Grammars	Grammars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	7	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	239	8	Depen-	Depen-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	9	dency	dency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	239	10	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	239	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	240	1	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	240	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	240	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	240	4	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	240	5	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	6	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	240	7	Empirical	Empirical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	8	Methods	Methods	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	240	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	240	10	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	11	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	12	Process-	Process-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	13	ing	ing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	240	15	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	240	16	1204–1213	1204–1213	CD
work_j4jyyr4gnrbelgfykpcumrgui4	240	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	240	18	October	October	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	240	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	241	1	Prachya	Prachya	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	241	2	Boonkwan	Boonkwan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	241	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	241	4	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	241	5	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	241	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	242	1	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	242	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	243	1	Gram-	Gram-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	2	mar	mar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	3	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	4	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	243	5	Text	Text	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	6	Using	Using	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	7	Small	Small	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	8	Syntactic	Syntactic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	9	Proto-	Proto-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	243	10	types	type	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	243	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	244	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	244	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	244	4	5th	5th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	244	5	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	6	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	7	Con-	Con-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	8	ference	ference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	244	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	244	10	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	11	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	12	Processing	Processing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	244	14	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	244	15	438	438	CD
work_j4jyyr4gnrbelgfykpcumrgui4	244	16	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	244	17	446	446	CD
work_j4jyyr4gnrbelgfykpcumrgui4	244	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	244	19	Chiang	Chiang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	20	Mai	Mai	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	244	22	Thailand	Thailand	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	244	24	November	November	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	244	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	245	1	Sabine	Sabine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	245	2	Buchholz	Buchholz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	245	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	245	4	Erwin	Erwin	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	245	5	Marsi	Marsi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	245	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	246	1	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	246	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	247	1	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	247	2	-	-	:
work_j4jyyr4gnrbelgfykpcumrgui4	247	3	X	x	NN
work_j4jyyr4gnrbelgfykpcumrgui4	247	4	Shared	share	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	247	5	Task	Task	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	247	6	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	247	7	Multilingual	multilingual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	247	8	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	247	9	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	247	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	248	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	248	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	248	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	248	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	248	5	10th	10th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	248	6	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	7	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	248	8	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	9	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	10	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	11	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	248	13	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	14	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	248	15	X	X	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	16	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	248	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	248	18	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	248	19	149	149	CD
work_j4jyyr4gnrbelgfykpcumrgui4	248	20	–	–	SYM
work_j4jyyr4gnrbelgfykpcumrgui4	248	21	164	164	CD
work_j4jyyr4gnrbelgfykpcumrgui4	248	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	248	23	New	New	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	24	York	York	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	25	City	City	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	248	27	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	248	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	249	1	Glenn	Glenn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	249	2	Carroll	Carroll	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	249	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	249	4	Eugene	Eugene	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	249	5	Charniak	Charniak	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	249	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	250	1	1992	1992	CD
work_j4jyyr4gnrbelgfykpcumrgui4	250	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	251	1	Two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	251	2	Exper-	Exper-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	3	iments	iment	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	251	4	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	251	5	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	6	Probabilistic	Probabilistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	7	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	8	Gram-	Gram-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	9	mars	mar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	251	10	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	251	11	Corpora	Corpora	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	251	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	252	1	Working	Working	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	252	2	Notes	Notes	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	252	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	252	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	252	5	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	252	6	Statistically	statistically	RB
work_j4jyyr4gnrbelgfykpcumrgui4	252	7	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	252	8	Based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	252	9	NLP	NLP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	252	10	Techniques	Techniques	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	252	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	252	12	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	252	13	1–13	1–13	VBP
work_j4jyyr4gnrbelgfykpcumrgui4	252	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	253	1	Eugene	Eugene	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	253	2	Charniak	Charniak	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	253	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	254	1	1993	1993	CD
work_j4jyyr4gnrbelgfykpcumrgui4	254	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	255	1	Statistical	statistical	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	255	2	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	255	3	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	255	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	256	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	256	2	MIT	MIT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	256	3	Press	Press	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	256	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	256	5	Cambridge	Cambridge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	256	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	256	7	Massachusetts	Massachusetts	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	256	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	257	1	Stephen	Stephen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	257	2	Clark	Clark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	257	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	257	4	James	James	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	257	5	R	R	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	257	6	Curran	Curran	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	257	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	258	1	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	258	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	259	1	Formalism-	Formalism-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	2	Independent	Independent	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	3	Parser	Parser	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	4	Evaluation	Evaluation	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	5	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	259	6	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	259	8	Dep-	Dep-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	9	Bank	Bank	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	259	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	260	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	260	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	260	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	260	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	260	5	45th	45th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	260	6	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	7	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	260	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	260	10	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	260	12	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	13	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	260	15	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	260	16	248–255	248–255	CD
work_j4jyyr4gnrbelgfykpcumrgui4	260	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	260	18	Prague	Prague	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	260	20	Czech	Czech	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	21	Republic	Republic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	260	23	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	260	24	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	261	1	Alex	Alex	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	261	2	Clark	Clark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	261	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	262	1	2001	2001	CD
work_j4jyyr4gnrbelgfykpcumrgui4	262	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	263	1	Unsupervised	Unsupervised	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	263	2	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	263	3	Acquisition	acquisition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	263	4	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	263	5	Theory	Theory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	263	6	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	263	7	Practice	Practice	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	263	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	264	1	Ph.D.	Ph.D.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	264	2	thesis	thesis	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	264	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	264	4	University	University	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	264	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	264	6	Sus-	Sus-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	264	7	sex	sex	NN
work_j4jyyr4gnrbelgfykpcumrgui4	264	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	264	9	September	September	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	264	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	265	1	Shay	Shay	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	2	B	B	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	3	Cohen	Cohen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	265	5	Noah	Noah	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	6	A	A	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	7	Smith	Smith	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	265	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	266	1	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	266	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	267	1	Variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	267	2	Inference	inference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	267	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	267	4	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	267	5	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	267	6	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	267	7	Prior	Prior	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	267	8	Knowl-	Knowl-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	267	9	edge	edge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	267	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	268	1	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	268	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	268	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	268	4	ACL	ACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	268	5	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	268	6	IJCNLP	IJCNLP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	268	7	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	268	8	Confer-	Confer-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	268	9	ence	ence	NN
work_j4jyyr4gnrbelgfykpcumrgui4	268	10	Short	Short	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	268	11	Papers	Papers	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	268	12	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	268	13	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	268	14	1–4	1–4	CD
work_j4jyyr4gnrbelgfykpcumrgui4	268	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	269	1	Shay	Shay	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	2	B	B	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	3	Cohen	Cohen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	269	5	Noah	Noah	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	6	A	A	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	7	Smith	Smith	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	269	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	270	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	270	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	271	1	Covariance	covariance	NN
work_j4jyyr4gnrbelgfykpcumrgui4	271	2	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	271	3	Unsupervised	Unsupervised	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	271	4	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	271	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	271	6	Probabilistic	Probabilistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	271	7	Grammars	Grammars	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	271	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	272	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	272	2	Journal	Journal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	272	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	272	4	Machine	Machine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	272	5	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	272	6	Research	Research	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	272	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	272	8	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	272	9	3117–3151	3117–3151	CD
work_j4jyyr4gnrbelgfykpcumrgui4	272	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	272	11	November	November	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	272	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	273	1	Trevor	Trevor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	2	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	273	4	Phil	Phil	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	5	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	273	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	273	8	Sharon	Sharon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	9	Goldwater	Goldwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	273	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	274	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	274	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	275	1	Inducing	induce	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	275	2	Tree	Tree	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	275	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	275	4	Substitution	Substitution	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	275	5	Grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	275	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	276	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	276	2	Journal	Journal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	276	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	276	4	Machine	Machine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	276	5	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	276	6	Research	Research	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	276	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	276	8	11:3053	11:3053	CD
work_j4jyyr4gnrbelgfykpcumrgui4	276	9	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	276	10	3096	3096	CD
work_j4jyyr4gnrbelgfykpcumrgui4	276	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	276	12	November	November	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	276	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	277	1	Michael	Michael	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	277	2	Collins	Collins	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	277	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	278	1	2003	2003	CD
work_j4jyyr4gnrbelgfykpcumrgui4	278	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	279	1	Head	Head	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	279	3	Driven	Driven	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	4	Statistical	Statistical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	5	Mod-	Mod-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	6	els	el	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	279	7	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	279	8	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	9	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	10	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	279	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	280	1	Computational	computational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	280	2	Lin-	Lin-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	280	3	guistics	guistic	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	280	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	280	5	29(4):589–637	29(4):589–637	CD
work_j4jyyr4gnrbelgfykpcumrgui4	280	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	280	7	December	December	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	280	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	281	1	Gregory	Gregory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	281	2	Druck	Druck	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	281	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	281	4	Gideon	Gideon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	281	5	Mann	Mann	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	281	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	281	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	281	8	Andrew	Andrew	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	281	9	McCal-	McCal-	:
work_j4jyyr4gnrbelgfykpcumrgui4	281	10	lum	lum	NN
work_j4jyyr4gnrbelgfykpcumrgui4	281	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	282	1	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	282	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	283	1	Semi	semi	RB
work_j4jyyr4gnrbelgfykpcumrgui4	283	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	283	3	supervised	supervised	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	283	4	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	283	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	283	6	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	283	7	Parsers	Parsers	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	283	8	using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	283	9	Generalized	Generalized	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	283	10	Expectation	Expectation	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	283	11	Criteria	Criteria	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	283	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	284	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	5	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	6	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	9	47th	47th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	284	10	An-	an-	CC
work_j4jyyr4gnrbelgfykpcumrgui4	284	11	nual	nual	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	284	12	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	15	ACL	ACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	284	17	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	18	4th	4th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	284	19	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	20	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	21	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	22	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	23	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	24	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	25	Processing	Processing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	26	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	284	27	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	28	AFNLP	AFNLP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	284	30	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	284	31	360–368	360–368	CD
work_j4jyyr4gnrbelgfykpcumrgui4	284	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	284	33	Suntec	Suntec	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	34	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	284	35	Singapore	Singapore	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	284	36	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	284	37	Au-	au-	DT
work_j4jyyr4gnrbelgfykpcumrgui4	284	38	gust	gust	NN
work_j4jyyr4gnrbelgfykpcumrgui4	284	39	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	285	1	Jason	Jason	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	285	2	Eisner	Eisner	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	285	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	286	1	1996	1996	CD
work_j4jyyr4gnrbelgfykpcumrgui4	286	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	287	1	Efficient	efficient	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	287	2	Normal	normal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	287	3	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	287	4	Form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	287	5	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	287	6	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	287	7	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	287	8	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	287	9	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	287	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	288	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	288	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	288	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	288	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	288	5	34th	34th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	288	6	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	7	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	288	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	288	10	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	11	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	288	12	Com-	Com-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	13	putational	putational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	288	14	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	288	16	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	288	17	79–86	79–86	CD
work_j4jyyr4gnrbelgfykpcumrgui4	288	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	288	19	Santa	Santa	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	20	Cruz	Cruz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	288	22	Cali-	cali-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	288	23	fornia	fornia	NN
work_j4jyyr4gnrbelgfykpcumrgui4	288	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	288	25	USA	USA	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	288	27	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	288	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	289	1	Douwe	Douwe	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	2	Gelling	gel	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	289	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	289	4	Trevor	Trevor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	5	Cohn	Cohn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	289	7	Phil	Phil	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	8	Blunsom	Blunsom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	289	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	289	11	João	João	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	12	V	V	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	13	Graca	Graca	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	289	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	290	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	290	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	291	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	291	2	PASCAL	pascal	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	291	3	Challenge	Challenge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	291	4	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	291	5	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	291	6	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	291	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	292	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	292	2	NAACL	NAACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	3	HLT	HLT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	4	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	5	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	292	6	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	292	8	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	9	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	292	11	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	292	12	64–80	64–80	CD
work_j4jyyr4gnrbelgfykpcumrgui4	292	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	292	14	Montréal	Montréal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	292	16	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	292	18	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	292	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	293	1	Jennifer	Jennifer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	2	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	293	4	Kuzman	Kuzman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	5	Ganchev	Ganchev	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	293	7	João	João	.
work_j4jyyr4gnrbelgfykpcumrgui4	293	8	V	V	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	9	Graca	Graca	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	293	11	Fernando	Fernando	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	12	Pereira	Pereira	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	293	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	293	15	Ben	Ben	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	16	Taskar	Taskar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	293	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	294	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	294	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	295	1	Sparsity	sparsity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	295	2	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	295	3	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	295	4	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	295	5	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	295	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	296	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	296	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	296	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	296	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	296	5	48th	48th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	296	6	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	7	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	296	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	296	10	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	11	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	296	12	Com-	Com-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	13	putational	putational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	296	14	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	296	16	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	296	17	194–199	194–199	CD
work_j4jyyr4gnrbelgfykpcumrgui4	296	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	296	19	Uppsala	Uppsala	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	296	21	Swe-	swe-	XX
work_j4jyyr4gnrbelgfykpcumrgui4	296	22	den	den	NN
work_j4jyyr4gnrbelgfykpcumrgui4	296	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	296	24	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	296	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	297	1	Jennifer	Jennifer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	2	Gillenwater	Gillenwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	297	4	Kuzman	Kuzman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	5	Ganchev	Ganchev	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	297	7	João	João	.
work_j4jyyr4gnrbelgfykpcumrgui4	297	8	V	V	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	9	Graca	Graca	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	297	11	Fernando	Fernando	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	12	Pereira	Pereira	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	297	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	297	15	Ben	Ben	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	16	Taskar	Taskar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	297	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	298	1	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	298	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	299	1	Posterior	Posterior	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	299	2	Sparsity	Sparsity	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	299	3	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	299	4	Unsupervised	Unsupervised	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	299	5	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	299	6	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	299	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	300	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	300	2	Journal	Journal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	300	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	300	4	Machine	Machine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	300	5	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	300	6	Research	Research	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	300	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	300	8	12:455–490	12:455–490	CD
work_j4jyyr4gnrbelgfykpcumrgui4	300	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	300	10	February	February	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	300	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	301	1	Yoav	Yoav	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	301	2	Goldberg	Goldberg	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	301	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	302	1	2011	2011	CD
work_j4jyyr4gnrbelgfykpcumrgui4	302	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	303	1	Automatic	automatic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	303	2	Syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	303	3	Processing	Processing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	303	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	303	5	Modern	Modern	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	303	6	Hebrew	Hebrew	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	303	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	304	1	Ph.D.	Ph.D.	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	2	thesis	thesis	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	304	4	Ben	Ben	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	5	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	304	6	Gurion	Gurion	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	7	University	University	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	304	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	304	10	Negev	Negev	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	304	12	November	November	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	304	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	305	1	William	William	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	2	P	P	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	3	Headden	Headden	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	4	III	III	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	305	6	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	7	Johnson	Johnson	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	305	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	305	10	David	David	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	11	Mc-	Mc-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	12	Closky	Closky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	305	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	306	1	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	306	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	307	1	Improving	improve	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	307	2	Unsupervised	Unsupervised	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	307	3	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	307	4	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	307	5	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	307	6	Richer	Richer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	307	7	Contexts	Contexts	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	307	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	307	9	Smoothing	smooth	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	307	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	308	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	308	2	Pro-	Pro-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	3	ceedings	ceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	308	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	308	5	Human	Human	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	6	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	7	Technologies	Technologies	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	308	8	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	308	9	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	308	10	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	308	11	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	12	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	13	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	308	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	308	15	North	north	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	308	16	American	American	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	17	Chapter	Chapter	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	18	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	308	19	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	308	20	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	21	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	308	22	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	23	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	308	25	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	308	26	101–109	101–109	CD
work_j4jyyr4gnrbelgfykpcumrgui4	308	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	308	28	Boulder	Boulder	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	29	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	308	30	Colorado	Colorado	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	31	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	308	32	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	308	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	309	1	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	309	2	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	309	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	309	4	Yonatan	Yonatan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	309	5	Bisk	Bisk	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	309	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	310	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	310	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	311	1	Normal-	Normal-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	311	2	form	form	NN
work_j4jyyr4gnrbelgfykpcumrgui4	311	3	parsing	parse	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	311	4	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	311	5	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	311	6	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	311	7	Grammars	Grammars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	311	8	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	311	9	generalized	generalized	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	311	10	composition	composition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	311	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	311	12	type	type	NN
work_j4jyyr4gnrbelgfykpcumrgui4	311	13	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	311	14	raising	raising	NN
work_j4jyyr4gnrbelgfykpcumrgui4	311	15	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	312	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	312	2	85	85	CD
work_j4jyyr4gnrbelgfykpcumrgui4	312	3	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	312	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	312	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	312	6	23rd	23rd	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	312	7	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	8	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	312	10	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	11	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	312	13	Coling	Coling	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	14	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	312	15	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	312	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	312	17	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	312	18	465	465	CD
work_j4jyyr4gnrbelgfykpcumrgui4	312	19	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	312	20	473	473	CD
work_j4jyyr4gnrbelgfykpcumrgui4	312	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	312	22	Beijing	Beijing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	312	24	China	China	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	312	26	August	August	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	312	27	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	313	1	Coling	Coling	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	313	2	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	313	3	Organizing	Organizing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	313	4	Committee	Committee	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	313	5	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	314	1	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	314	2	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	314	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	314	4	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	314	5	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	314	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	315	1	2002	2002	CD
work_j4jyyr4gnrbelgfykpcumrgui4	315	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	316	1	Gener-	Gener-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	2	ative	ative	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	316	3	Models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	316	4	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	316	5	Statistical	Statistical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	6	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	7	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	316	8	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	9	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	10	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	316	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	317	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	317	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	317	4	40th	40th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	317	5	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	6	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	317	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	317	9	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	10	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	317	11	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	12	Lin-	Lin-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	13	guistics	guistic	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	317	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	317	15	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	317	16	335–342	335–342	CD
work_j4jyyr4gnrbelgfykpcumrgui4	317	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	317	18	Philadelphia	Philadelphia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	317	20	Pennsylvania	Pennsylvania	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	317	22	USA	USA	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	317	24	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	317	25	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	318	1	Julia	Julia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	318	2	Hockenmaier	Hockenmaier	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	318	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	318	4	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	318	5	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	318	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	319	1	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	319	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	320	1	CCG-	CCG-	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	320	2	bank	bank	NN
work_j4jyyr4gnrbelgfykpcumrgui4	320	3	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	320	4	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	320	5	Corpus	Corpus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	320	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	320	7	CCG	CCG	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	320	8	Derivations	Derivations	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	320	9	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	320	10	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	320	11	Structures	Structures	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	320	12	Extracted	extract	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	320	13	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	320	14	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	320	15	Penn	Penn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	320	16	Treebank	Treebank	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	320	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	321	1	Com-	Com-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	321	2	putational	putational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	321	3	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	321	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	321	5	33(3):355–396	33(3):355–396	CD
work_j4jyyr4gnrbelgfykpcumrgui4	321	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	321	7	September	September	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	321	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	322	1	Yun	Yun	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	2	Huang	Huang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	322	4	Min	Min	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	5	Zhang	Zhang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	322	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	322	8	Chew	Chew	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	9	Lim	Lim	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	10	Tan	Tan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	322	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	323	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	323	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	324	1	Improved	Improved	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	2	Combinatory	Combinatory	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	3	Categorial	Categorial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	4	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	5	Induc-	Induc-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	6	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	324	7	with	with	IN
work_j4jyyr4gnrbelgfykpcumrgui4	324	8	Boundary	Boundary	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	9	Words	Words	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	324	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	324	11	Bayesian	Bayesian	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	12	Inference	Inference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	324	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	325	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	325	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	325	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	325	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	325	5	24rd	24rd	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	325	6	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	7	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	8	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	325	9	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	10	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	11	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	325	12	Coling	Coling	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	13	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	325	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	325	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	325	16	Mumbai	Mumbai	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	325	18	India	India	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	325	20	December	December	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	325	21	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	326	1	Ray	Ray	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	326	2	Jackendoff	Jackendoff	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	326	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	327	1	1977	1977	CD
work_j4jyyr4gnrbelgfykpcumrgui4	327	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	328	1	X	x	LS
work_j4jyyr4gnrbelgfykpcumrgui4	328	2	-	-	:
work_j4jyyr4gnrbelgfykpcumrgui4	328	3	Bar	bar	NN
work_j4jyyr4gnrbelgfykpcumrgui4	328	4	Syntax	syntax	NN
work_j4jyyr4gnrbelgfykpcumrgui4	328	5	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	328	6	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	328	7	Study	Study	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	328	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	328	9	Phrase	Phrase	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	328	10	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	328	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	329	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	329	2	MIT	MIT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	329	3	Press	Press	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	329	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	330	1	Dan	Dan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	330	2	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	330	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	330	4	Christopher	Christopher	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	330	5	D	D	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	330	6	Manning	Manning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	330	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	331	1	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	331	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	332	1	Corpus-	Corpus-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	2	Based	Based	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	3	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	332	5	Syntactic	Syntactic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	6	Structure	structure	NN
work_j4jyyr4gnrbelgfykpcumrgui4	332	7	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	332	8	Models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	332	9	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	332	10	De-	De-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	11	pendency	pendency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	332	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	332	13	Constituency	Constituency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	332	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	333	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	333	2	Proceedings	Proceedings	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	333	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	333	5	42nd	42nd	NN
work_j4jyyr4gnrbelgfykpcumrgui4	333	6	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	333	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	333	9	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	10	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	333	11	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	12	Linguistics	Linguistics	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	333	13	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	333	14	ACL’04	ACL’04	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	15	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	333	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	333	17	Main	Main	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	18	Volume	Volume	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	333	20	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	333	21	478–485	478–485	CD
work_j4jyyr4gnrbelgfykpcumrgui4	333	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	333	23	Barcelona	Barcelona	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	333	25	Spain	Spain	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	333	27	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	333	28	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	334	1	Kenichi	Kenichi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	334	2	Kurihara	Kurihara	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	334	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	334	4	Taisuke	Taisuke	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	334	5	Sato	Sato	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	334	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	335	1	2004	2004	CD
work_j4jyyr4gnrbelgfykpcumrgui4	335	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	336	1	An	an	DT
work_j4jyyr4gnrbelgfykpcumrgui4	336	2	Appli-	Appli-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	3	cation	cation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	336	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	336	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	336	6	Variational	Variational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	7	Bayesian	Bayesian	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	8	Approach	Approach	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	9	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	336	10	Prob-	prob-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	336	11	abilistic	abilistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	336	12	Context	Context	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	13	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	336	14	Free	Free	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	336	15	Grammars	Grammars	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	336	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	337	1	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	2	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	3	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	4	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	337	5	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	6	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	7	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	8	Process-	Process-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	9	ing	ing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	10	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	11	Beyond	beyond	IN
work_j4jyyr4gnrbelgfykpcumrgui4	337	12	Shallow	Shallow	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	13	Analyses	Analyses	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	337	15	March	March	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	337	16	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	338	1	Kenichi	Kenichi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	2	Kurihara	Kurihara	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	338	4	Max	Max	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	5	Welling	Welling	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	338	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	338	8	Yee	Yee	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	9	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	338	10	Whye	Whye	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	11	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	338	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	339	1	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	339	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	340	1	Collapsed	collapse	VBD
work_j4jyyr4gnrbelgfykpcumrgui4	340	2	Variational	variational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	340	3	Dirichlet	dirichlet	NN
work_j4jyyr4gnrbelgfykpcumrgui4	340	4	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	340	5	Mix-	Mix-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	340	6	ture	ture	NN
work_j4jyyr4gnrbelgfykpcumrgui4	340	7	Models	model	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	340	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	341	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	341	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	341	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	341	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	341	5	20th	20th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	341	6	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	7	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	8	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	341	10	Artificial	Artificial	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	11	Intelligence	Intelligence	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	12	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	341	13	IJCAI07	IJCAI07	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	14	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	341	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	341	16	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	341	17	2796–2801	2796–2801	CD
work_j4jyyr4gnrbelgfykpcumrgui4	341	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	341	19	Hyderabad	Hyderabad	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	341	21	India	India	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	341	23	January	January	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	341	24	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	342	1	Tom	Tom	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	2	Kwiatkowski	Kwiatkowski	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	342	4	Sharon	Sharon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	5	Goldwater	Goldwater	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	342	7	Luke	Luke	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	8	Zettlemoyer	Zettlemoyer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	342	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	342	11	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	12	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	342	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	343	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	343	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	344	1	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	344	2	probabilistic	probabilistic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	344	3	model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	344	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	344	5	syntactic	syntactic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	344	6	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	344	7	semantic	semantic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	344	8	acquisition	acquisition	NN
work_j4jyyr4gnrbelgfykpcumrgui4	344	9	from	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	344	10	child	child	NN
work_j4jyyr4gnrbelgfykpcumrgui4	344	11	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	344	12	directed	direct	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	344	13	utterances	utterance	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	344	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	344	15	their	-PRON-	PRP$
work_j4jyyr4gnrbelgfykpcumrgui4	344	16	meanings	meaning	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	344	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	345	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	345	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	345	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	345	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	345	5	13th	13th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	345	6	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	345	8	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	345	9	European	European	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	10	Chapter	Chapter	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	11	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	345	12	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	345	13	As-	As-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	14	sociation	sociation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	345	15	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	345	16	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	17	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	345	19	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	345	20	234	234	CD
work_j4jyyr4gnrbelgfykpcumrgui4	345	21	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	345	22	244	244	CD
work_j4jyyr4gnrbelgfykpcumrgui4	345	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	345	24	Avignon	Avignon	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	345	26	France	France	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	345	28	April	April	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	345	29	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	346	1	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	346	2	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	346	3	Compu-	Compu-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	346	4	tational	tational	NN
work_j4jyyr4gnrbelgfykpcumrgui4	346	5	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	346	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	347	1	Karim	Karim	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	347	2	Lari	Lari	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	347	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	347	4	Steve	Steve	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	347	5	J	J	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	347	6	Young	Young	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	347	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	348	1	1991	1991	CD
work_j4jyyr4gnrbelgfykpcumrgui4	348	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	349	1	Applications	application	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	349	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	349	3	stochastic	stochastic	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	349	4	context	context	NN
work_j4jyyr4gnrbelgfykpcumrgui4	349	5	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	349	6	free	free	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	349	7	grammars	grammar	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	349	8	using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	349	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	349	10	Inside-	Inside-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	349	11	Outside	Outside	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	349	12	algorithm	algorithm	NN
work_j4jyyr4gnrbelgfykpcumrgui4	349	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	350	1	Computer	computer	NN
work_j4jyyr4gnrbelgfykpcumrgui4	350	2	speech	speech	NN
work_j4jyyr4gnrbelgfykpcumrgui4	350	3	&	&	CC
work_j4jyyr4gnrbelgfykpcumrgui4	350	4	language	language	NN
work_j4jyyr4gnrbelgfykpcumrgui4	350	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	350	6	5(3):237–257	5(3):237–257	CD
work_j4jyyr4gnrbelgfykpcumrgui4	350	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	350	8	January	January	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	350	9	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	351	1	Percy	Percy	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	2	Liang	Liang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	351	4	Slav	Slav	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	5	Petrov	Petrov	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	351	7	Michael	Michael	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	8	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	9	Jordan	Jordan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	351	11	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	351	12	Dan	Dan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	13	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	351	14	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	352	1	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	352	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	353	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	353	2	Infinite	infinite	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	353	3	PCFG	pcfg	NN
work_j4jyyr4gnrbelgfykpcumrgui4	353	4	Using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	353	5	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	353	6	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	353	7	Processes	Processes	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	353	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	354	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	354	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	354	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	354	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	354	5	2007	2007	CD
work_j4jyyr4gnrbelgfykpcumrgui4	354	6	Joint	Joint	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	7	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	8	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	354	9	Empirical	Empirical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	10	Methods	Methods	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	354	11	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	354	12	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	13	Lan-	Lan-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	14	guage	guage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	354	15	Processing	Processing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	16	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	354	17	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	18	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	19	Lan-	Lan-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	20	guage	guage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	354	21	Learning	learn	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	354	22	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	354	23	EMNLP	EMNLP	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	24	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	354	25	CoNLL	CoNLL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	26	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	354	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	354	28	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	354	29	688–697	688–697	CD
work_j4jyyr4gnrbelgfykpcumrgui4	354	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	354	31	Prague	Prague	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	354	33	Czech	Czech	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	34	Republic	Republic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	354	35	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	355	1	Percy	Percy	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	2	Liang	Liang	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	355	4	Michael	Michael	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	5	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	6	Jordan	Jordan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	355	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	355	9	Dan	Dan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	10	Klein	Klein	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	355	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	356	1	2009	2009	CD
work_j4jyyr4gnrbelgfykpcumrgui4	356	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	357	1	Probabilistic	Probabilistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	357	2	Grammars	Grammars	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	357	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	357	4	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	357	5	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	357	6	Processes	Processes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	357	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	358	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	358	2	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	358	3	Oxford	Oxford	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	358	4	Handbook	Handbook	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	358	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	358	6	Applied	Applied	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	358	7	Bayesian	Bayesian	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	358	8	Analysis	Analysis	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	358	9	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	359	1	Oxford	Oxford	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	359	2	University	University	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	359	3	Press	Press	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	359	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	360	1	Mitchell	Mitchell	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	2	P	P	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	3	Marcus	Marcus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	360	5	Beatrice	Beatrice	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	6	Santorini	Santorini	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	360	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	360	9	Mary	Mary	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	10	Ann	Ann	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	11	Marcinkiewicz	Marcinkiewicz	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	360	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	361	1	1993	1993	CD
work_j4jyyr4gnrbelgfykpcumrgui4	361	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	362	1	Building	build	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	362	2	a	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	362	3	Large	large	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	362	4	Annotated	annotate	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	362	5	Corpus	Corpus	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	362	6	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	362	7	English	English	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	362	8	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	362	9	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	362	10	Penn	Penn	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	362	11	Treebank	Treebank	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	362	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	363	1	Computa-	Computa-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	363	2	tional	tional	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	363	3	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	363	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	363	5	19(2):313–330	19(2):313–330	CD
work_j4jyyr4gnrbelgfykpcumrgui4	363	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	363	7	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	363	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	364	1	David	David	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	364	2	Marecek	Marecek	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	364	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	364	4	Zdenek	Zdenek	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	364	5	Zabokrtsky	Zabokrtsky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	364	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	365	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	365	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	366	1	Unsu-	Unsu-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	366	2	pervised	pervised	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	366	3	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	366	4	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	366	5	using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	366	6	Reducibility	Reducibility	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	366	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	366	8	Fertility	Fertility	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	366	9	features	feature	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	366	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	367	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	367	2	NAACL	NAACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	3	HLT	HLT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	4	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	5	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	367	6	Induc-	Induc-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	7	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	367	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	367	9	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	10	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	367	12	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	367	13	84–89	84–89	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	367	15	Montréal	Montréal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	367	17	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	367	19	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	367	20	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	368	1	Tahira	Tahira	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	2	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	368	4	Harr	Harr	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	5	Chen	Chen	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	368	7	Regina	Regina	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	8	Barzilay	Barzilay	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	368	10	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	368	11	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	12	Johnson	Johnson	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	368	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	369	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	369	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	370	1	Using	use	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	370	2	Universal	Universal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	3	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	4	Knowl-	Knowl-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	5	edge	edge	NN
work_j4jyyr4gnrbelgfykpcumrgui4	370	6	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	370	7	Guide	Guide	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	8	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	9	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	370	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	371	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	371	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	371	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	371	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	371	5	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	371	6	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	7	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	371	8	Empirical	Empirical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	9	Methods	Methods	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	371	10	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	371	11	Natural	Natural	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	12	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	13	Processing	Processing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	14	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	371	15	pages	page	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	371	16	1234–1244	1234–1244	CD
work_j4jyyr4gnrbelgfykpcumrgui4	371	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	371	18	Cambridge	Cambridge	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	19	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	371	20	MA	MA	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	21	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	371	22	October	October	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	371	23	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	372	1	Tahira	Tahira	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	2	Naseem	Naseem	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	372	4	Regina	Regina	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	5	Barzilay	Barzilay	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	372	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	372	8	Amir	Amir	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	9	Globerson	Globerson	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	372	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	373	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	373	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	374	1	Selective	Selective	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	374	2	Sharing	Sharing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	374	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	374	4	Multilingual	Multilingual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	374	5	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	374	6	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	374	7	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	375	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	375	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	375	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	375	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	375	5	50th	50th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	375	6	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	7	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	375	9	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	375	10	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	11	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	375	12	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	13	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	14	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	375	15	Vol-	Vol-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	16	ume	ume	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	17	1	1	CD
work_j4jyyr4gnrbelgfykpcumrgui4	375	18	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	375	19	Long	Long	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	20	Papers	Papers	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	375	21	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	375	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	375	23	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	375	24	629–637	629–637	CD
work_j4jyyr4gnrbelgfykpcumrgui4	375	25	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	375	26	Jeju	Jeju	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	375	28	Republic	Republic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	29	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	375	30	Korea	Korea	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	31	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	375	32	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	375	33	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	376	1	Joakim	Joakim	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	376	2	Nivre	Nivre	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	376	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	377	1	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	377	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	378	1	Inductive	Inductive	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	378	2	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	378	3	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	378	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	379	1	Springer	Springer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	379	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	380	1	Slav	Slav	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	2	Petrov	Petrov	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	3	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	380	4	Dipanjan	Dipanjan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	5	Das	Das	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	6	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	380	7	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	380	8	Ryan	Ryan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	9	McDonald	McDonald	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	380	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	381	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	381	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	382	1	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	382	2	Universal	Universal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	382	3	Part	Part	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	382	4	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	382	5	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	382	6	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	382	7	Speech	Speech	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	382	8	Tagset	Tagset	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	382	9	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	383	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	383	2	Proceedings	proceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	383	3	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	383	4	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	383	5	8th	8th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	383	6	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	7	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	8	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	383	9	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	10	Re-	Re-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	11	sources	source	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	383	12	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	383	13	Evaluation	Evaluation	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	14	(	(	-LRB-
work_j4jyyr4gnrbelgfykpcumrgui4	383	15	LREC-2012	LREC-2012	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	16	)	)	-RRB-
work_j4jyyr4gnrbelgfykpcumrgui4	383	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	383	18	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	383	19	2089	2089	CD
work_j4jyyr4gnrbelgfykpcumrgui4	383	20	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	383	21	2096	2096	CD
work_j4jyyr4gnrbelgfykpcumrgui4	383	22	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	383	23	Istanbul	Istanbul	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	383	25	Turkey	Turkey	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	383	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	383	27	May.	May	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	384	1	Anders	Anders	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	384	2	Søgaard	Søgaard	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	384	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	385	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	385	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	386	1	Two	two	CD
work_j4jyyr4gnrbelgfykpcumrgui4	386	2	baselines	baseline	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	386	3	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	386	4	unsuper-	unsuper-	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	386	5	vised	vised	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	386	6	dependency	dependency	NN
work_j4jyyr4gnrbelgfykpcumrgui4	386	7	parsing	parsing	NN
work_j4jyyr4gnrbelgfykpcumrgui4	386	8	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	387	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	387	2	NAACL	NAACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	3	HLT	HLT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	4	Work-	work-	NN
work_j4jyyr4gnrbelgfykpcumrgui4	387	5	shop	shop	NN
work_j4jyyr4gnrbelgfykpcumrgui4	387	6	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	387	7	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	8	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	387	9	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	10	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	11	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	387	12	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	387	13	81	81	CD
work_j4jyyr4gnrbelgfykpcumrgui4	387	14	–	–	:
work_j4jyyr4gnrbelgfykpcumrgui4	387	15	83	83	CD
work_j4jyyr4gnrbelgfykpcumrgui4	387	16	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	387	17	Montréal	Montréal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	18	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	387	19	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	20	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	387	21	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	387	22	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	388	1	Valentin	Valentin	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	2	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	3	Spitkovsky	Spitkovsky	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	388	5	Hiyan	Hiyan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	6	Alshawi	Alshawi	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	388	8	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	388	9	Daniel	Daniel	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	10	Juraf-	Juraf-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	388	11	sky	sky	NN
work_j4jyyr4gnrbelgfykpcumrgui4	388	12	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	389	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	389	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	390	1	From	from	IN
work_j4jyyr4gnrbelgfykpcumrgui4	390	2	Baby	Baby	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	390	3	Steps	Steps	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	390	4	to	to	IN
work_j4jyyr4gnrbelgfykpcumrgui4	390	5	Leapfrog	Leapfrog	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	390	6	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	390	7	How	how	WRB
work_j4jyyr4gnrbelgfykpcumrgui4	390	8	“	"	``
work_j4jyyr4gnrbelgfykpcumrgui4	390	9	Less	Less	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	390	10	is	be	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	390	11	More	More	JJR
work_j4jyyr4gnrbelgfykpcumrgui4	390	12	”	"	''
work_j4jyyr4gnrbelgfykpcumrgui4	390	13	in	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	390	14	Unsupervised	Unsupervised	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	390	15	Dependency	Dependency	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	390	16	Parsing	Parsing	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	390	17	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	391	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	391	2	Hu-	Hu-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	3	man	man	NN
work_j4jyyr4gnrbelgfykpcumrgui4	391	4	Language	Language	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	5	Technologies	Technologies	NNPS
work_j4jyyr4gnrbelgfykpcumrgui4	391	6	:	:	:
work_j4jyyr4gnrbelgfykpcumrgui4	391	7	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	391	8	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	391	9	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	10	Con-	Con-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	11	ference	ference	NN
work_j4jyyr4gnrbelgfykpcumrgui4	391	12	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	391	13	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	391	14	North	north	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	391	15	American	american	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	391	16	Chapter	Chapter	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	391	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	391	19	Associ-	Associ-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	20	ation	ation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	391	21	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	391	22	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	23	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	24	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	391	25	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	391	26	751–759	751–759	CD
work_j4jyyr4gnrbelgfykpcumrgui4	391	27	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	391	28	Los	Los	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	29	Angeles	Angeles	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	391	31	California	California	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	32	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	391	33	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	391	34	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	392	1	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	392	2	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	392	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	393	1	1996	1996	CD
work_j4jyyr4gnrbelgfykpcumrgui4	393	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	394	1	Surface	Surface	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	394	2	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	394	3	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	394	4	Interpre-	Interpre-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	394	5	tation	tation	NN
work_j4jyyr4gnrbelgfykpcumrgui4	394	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	395	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	395	2	MIT	MIT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	395	3	Press	Press	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	395	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	395	5	January	January	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	395	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	396	1	Mark	Mark	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	396	2	Steedman	Steedman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	396	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	397	1	2000	2000	CD
work_j4jyyr4gnrbelgfykpcumrgui4	397	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	398	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	398	2	Syntactic	Syntactic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	398	3	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	398	4	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	399	1	The	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	399	2	MIT	MIT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	399	3	Press	Press	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	399	4	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	399	5	September	September	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	399	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	400	1	Yee	Yee	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	400	3	Whye	Whye	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	4	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	5	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	400	6	Michael	Michael	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	7	I	I	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	8	Jordan	Jordan	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	9	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	400	10	Matthew	Matthew	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	11	J	J	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	12	Beal	Beal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	400	14	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	400	15	David	David	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	16	M	M	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	17	Blei	Blei	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	400	18	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	401	1	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	401	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	402	1	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	402	2	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	402	3	Pro-	Pro-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	402	4	86	86	CD
work_j4jyyr4gnrbelgfykpcumrgui4	402	5	cesses	cesse	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	402	6	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	403	1	Journal	Journal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	403	2	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	403	3	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	403	4	American	American	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	403	5	Statistical	Statistical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	403	6	Associa-	Associa-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	403	7	tion	tion	NN
work_j4jyyr4gnrbelgfykpcumrgui4	403	8	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	403	9	101(476):1566–1581	101(476):1566–1581	CD
work_j4jyyr4gnrbelgfykpcumrgui4	403	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	404	1	Yee	Yee	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	404	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	404	3	Whye	Whye	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	404	4	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	404	5	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	405	1	2006	2006	CD
work_j4jyyr4gnrbelgfykpcumrgui4	405	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	406	1	A	a	DT
work_j4jyyr4gnrbelgfykpcumrgui4	406	2	Hierarchical	Hierarchical	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	3	Bayesian	Bayesian	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	4	Lan-	Lan-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	5	guage	guage	NN
work_j4jyyr4gnrbelgfykpcumrgui4	406	6	Model	model	NN
work_j4jyyr4gnrbelgfykpcumrgui4	406	7	based	base	VBN
work_j4jyyr4gnrbelgfykpcumrgui4	406	8	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	406	9	Pitman	Pitman	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	10	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	406	11	Yor	Yor	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	12	Processes	Processes	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	406	13	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	407	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	407	2	Pro-	Pro-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	3	ceedings	ceeding	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	407	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	407	5	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	407	6	21st	21st	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	407	7	International	International	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	8	Conference	Conference	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	9	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	407	10	Com-	Com-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	11	putational	putational	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	407	12	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	13	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	407	14	44th	44th	JJ
work_j4jyyr4gnrbelgfykpcumrgui4	407	15	Annual	Annual	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	16	Meeting	Meeting	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	17	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	407	18	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	407	19	Association	Association	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	20	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	407	21	Computational	Computational	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	22	Linguistics	Linguistics	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	23	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	407	24	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	407	25	985–992	985–992	CD
work_j4jyyr4gnrbelgfykpcumrgui4	407	26	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	407	27	Sydney	Sydney	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	28	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	407	29	Australia	Australia	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	30	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	407	31	July	July	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	407	32	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	408	1	Yee	Yee	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	408	2	-	-	HYPH
work_j4jyyr4gnrbelgfykpcumrgui4	408	3	Whye	Whye	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	408	4	Teh	Teh	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	408	5	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	409	1	2010	2010	CD
work_j4jyyr4gnrbelgfykpcumrgui4	409	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	410	1	Dirichlet	Dirichlet	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	410	2	Process	Process	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	410	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	411	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	411	2	Encyclope-	Encyclope-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	411	3	dia	dia	NN
work_j4jyyr4gnrbelgfykpcumrgui4	411	4	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	411	5	Machine	Machine	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	411	6	Learning	Learning	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	411	7	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	411	8	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	411	9	280–287	280–287	CD
work_j4jyyr4gnrbelgfykpcumrgui4	411	10	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	412	1	Springer	Springer	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	412	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	413	1	Kewei	Kewei	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	413	2	Tu	Tu	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	413	3	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	414	1	2012	2012	CD
work_j4jyyr4gnrbelgfykpcumrgui4	414	2	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	415	1	Combining	combine	VBG
work_j4jyyr4gnrbelgfykpcumrgui4	415	2	the	the	DT
work_j4jyyr4gnrbelgfykpcumrgui4	415	3	Sparsity	Sparsity	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	415	4	and	and	CC
work_j4jyyr4gnrbelgfykpcumrgui4	415	5	Unambi-	Unambi-	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	415	6	guity	guity	NN
work_j4jyyr4gnrbelgfykpcumrgui4	415	7	Biases	bias	NNS
work_j4jyyr4gnrbelgfykpcumrgui4	415	8	for	for	IN
work_j4jyyr4gnrbelgfykpcumrgui4	415	9	Grammar	Grammar	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	415	10	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	415	11	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	416	1	In	in	IN
work_j4jyyr4gnrbelgfykpcumrgui4	416	2	NAACL	NAACL	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	3	HLT	HLT	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	4	Workshop	Workshop	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	5	on	on	IN
work_j4jyyr4gnrbelgfykpcumrgui4	416	6	Induction	Induction	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	7	of	of	IN
work_j4jyyr4gnrbelgfykpcumrgui4	416	8	Linguistic	Linguistic	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	9	Structure	Structure	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	10	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	416	11	pages	page	VBZ
work_j4jyyr4gnrbelgfykpcumrgui4	416	12	105–110	105–110	CD
work_j4jyyr4gnrbelgfykpcumrgui4	416	13	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	416	14	Montréal	Montréal	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	15	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	416	16	Canada	Canada	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	17	,	,	,
work_j4jyyr4gnrbelgfykpcumrgui4	416	18	June	June	NNP
work_j4jyyr4gnrbelgfykpcumrgui4	416	19	.	.	.
work_j4jyyr4gnrbelgfykpcumrgui4	417	1	87	87	CD
work_j4jyyr4gnrbelgfykpcumrgui4	417	2	88	88	CD