id	sid	tid	token	lemma	pos
hansen-can-2020	1	1	Can	Can	MD
hansen-can-2020	1	2	a	a	DT
hansen-can-2020	1	3	Hammer	Hammer	NNP
hansen-can-2020	1	4	Categorize	Categorize	NNP
hansen-can-2020	1	5	Highly	highly	RB
hansen-can-2020	1	6	Technical	technical	JJ
hansen-can-2020	1	7	Articles	article	NNS
hansen-can-2020	1	8	?	?	.
hansen-can-2020	2	1	Samuel	Samuel	NNP
hansen-can-2020	2	2	Hansen	Hansen	NNP
hansen-can-2020	2	3	University	University	NNP
hansen-can-2020	2	4	of	of	IN
hansen-can-2020	2	5	Michigan	Michigan	NNP
hansen-can-2020	2	6	hansensm@umich.edu	hansensm@umich.edu	NNP
hansen-can-2020	2	7	When	when	WRB
hansen-can-2020	2	8	everything	everything	NN
hansen-can-2020	2	9	looks	look	VBZ
hansen-can-2020	2	10	like	like	IN
hansen-can-2020	2	11	a	a	DT
hansen-can-2020	2	12	nail	nail	NN
hansen-can-2020	2	13	...	...	.
hansen-can-2020	3	1	I	-PRON-	PRP
hansen-can-2020	3	2	was	be	VBD
hansen-can-2020	3	3	sure	sure	JJ
hansen-can-2020	3	4	I	-PRON-	PRP
hansen-can-2020	3	5	had	have	VBD
hansen-can-2020	3	6	the	the	DT
hansen-can-2020	3	7	most	most	RBS
hansen-can-2020	3	8	brilliant	brilliant	JJ
hansen-can-2020	3	9	research	research	NN
hansen-can-2020	3	10	project	project	NN
hansen-can-2020	3	11	idea	idea	NN
hansen-can-2020	3	12	for	for	IN
hansen-can-2020	3	13	my	-PRON-	PRP$
hansen-can-2020	3	14	course	course	NN
hansen-can-2020	3	15	in	in	IN
hansen-can-2020	3	16	Digital	Digital	NNP
hansen-can-2020	3	17	Scholarship	Scholarship	NNP
hansen-can-2020	3	18	techniques	technique	NNS
hansen-can-2020	3	19	.	.	.
hansen-can-2020	4	1	I	-PRON-	PRP
hansen-can-2020	4	2	would	would	MD
hansen-can-2020	4	3	use	use	VB
hansen-can-2020	4	4	the	the	DT
hansen-can-2020	4	5	Mathematical	Mathematical	NNP
hansen-can-2020	4	6	Subject	Subject	NNP
hansen-can-2020	4	7	Classification	Classification	NNP
hansen-can-2020	4	8	(	(	-LRB-
hansen-can-2020	4	9	MSC	MSC	NNP
hansen-can-2020	4	10	)	)	-RRB-
hansen-can-2020	4	11	values	value	NNS
hansen-can-2020	4	12	assigned	assign	VBN
hansen-can-2020	4	13	to	to	IN
hansen-can-2020	4	14	the	the	DT
hansen-can-2020	4	15	publications	publication	NNS
hansen-can-2020	4	16	in	in	IN
hansen-can-2020	4	17	MathSciNet[footnoteRef:0	MathSciNet[footnoteRef:0	NNP
hansen-can-2020	4	18	]	]	-RRB-
hansen-can-2020	4	19	to	to	TO
hansen-can-2020	4	20	create	create	VB
hansen-can-2020	4	21	a	a	DT
hansen-can-2020	4	22	temporal	temporal	JJ
hansen-can-2020	4	23	citation	citation	NN
hansen-can-2020	4	24	network	network	NN
hansen-can-2020	4	25	which	which	WDT
hansen-can-2020	4	26	would	would	MD
hansen-can-2020	4	27	allow	allow	VB
hansen-can-2020	4	28	me	-PRON-	PRP
hansen-can-2020	4	29	to	to	TO
hansen-can-2020	4	30	visualize	visualize	VB
hansen-can-2020	4	31	how	how	WRB
hansen-can-2020	4	32	new	new	JJ
hansen-can-2020	4	33	subfields	subfield	NNS
hansen-can-2020	4	34	were	be	VBD
hansen-can-2020	4	35	created	create	VBN
hansen-can-2020	4	36	and	and	CC
hansen-can-2020	4	37	perhaps	perhaps	RB
hansen-can-2020	4	38	even	even	RB
hansen-can-2020	4	39	predict	predict	VBP
hansen-can-2020	4	40	them	-PRON-	PRP
hansen-can-2020	4	41	while	while	IN
hansen-can-2020	4	42	they	-PRON-	PRP
hansen-can-2020	4	43	were	be	VBD
hansen-can-2020	4	44	still	still	RB
hansen-can-2020	4	45	in	in	IN
hansen-can-2020	4	46	their	-PRON-	PRP$
hansen-can-2020	4	47	infancy	infancy	NN
hansen-can-2020	4	48	.	.	.
hansen-can-2020	5	1	I	-PRON-	PRP
hansen-can-2020	5	2	thought	think	VBD
hansen-can-2020	5	3	it	-PRON-	PRP
hansen-can-2020	5	4	would	would	MD
hansen-can-2020	5	5	be	be	VB
hansen-can-2020	5	6	an	an	DT
hansen-can-2020	5	7	easy	easy	JJ
hansen-can-2020	5	8	enough	enough	JJ
hansen-can-2020	5	9	project	project	NN
hansen-can-2020	5	10	.	.	.
hansen-can-2020	6	1	I	-PRON-	PRP
hansen-can-2020	6	2	already	already	RB
hansen-can-2020	6	3	knew	know	VBD
hansen-can-2020	6	4	how	how	WRB
hansen-can-2020	6	5	to	to	TO
hansen-can-2020	6	6	analyze	analyze	VB
hansen-can-2020	6	7	network	network	NN
hansen-can-2020	6	8	data	datum	NNS
hansen-can-2020	6	9	and	and	CC
hansen-can-2020	6	10	the	the	DT
hansen-can-2020	6	11	data	datum	NNS
hansen-can-2020	6	12	I	-PRON-	PRP
hansen-can-2020	6	13	needed	need	VBD
hansen-can-2020	6	14	already	already	RB
hansen-can-2020	6	15	existed	exist	VBD
hansen-can-2020	6	16	,	,	,
hansen-can-2020	6	17	I	-PRON-	PRP
hansen-can-2020	6	18	just	just	RB
hansen-can-2020	6	19	had	have	VBD
hansen-can-2020	6	20	to	to	TO
hansen-can-2020	6	21	get	get	VB
hansen-can-2020	6	22	my	-PRON-	PRP$
hansen-can-2020	6	23	hands	hand	NNS
hansen-can-2020	6	24	on	on	IN
hansen-can-2020	6	25	it	-PRON-	PRP
hansen-can-2020	6	26	.	.	.
hansen-can-2020	7	1	I	-PRON-	PRP
hansen-can-2020	7	2	even	even	RB
hansen-can-2020	7	3	sold	sell	VBD
hansen-can-2020	7	4	a	a	DT
hansen-can-2020	7	5	couple	couple	NN
hansen-can-2020	7	6	of	of	IN
hansen-can-2020	7	7	my	-PRON-	PRP$
hansen-can-2020	7	8	fellow	fellow	JJ
hansen-can-2020	7	9	coursemates	coursemate	NNS
hansen-can-2020	7	10	on	on	IN
hansen-can-2020	7	11	the	the	DT
hansen-can-2020	7	12	idea	idea	NN
hansen-can-2020	7	13	and	and	CC
hansen-can-2020	7	14	they	-PRON-	PRP
hansen-can-2020	7	15	agreed	agree	VBD
hansen-can-2020	7	16	to	to	TO
hansen-can-2020	7	17	work	work	VB
hansen-can-2020	7	18	with	with	IN
hansen-can-2020	7	19	me	-PRON-	PRP
hansen-can-2020	7	20	.	.	.
hansen-can-2020	8	1	Of	of	RB
hansen-can-2020	8	2	course	course	RB
hansen-can-2020	8	3	nothing	nothing	NN
hansen-can-2020	8	4	is	be	VBZ
hansen-can-2020	8	5	as	as	RB
hansen-can-2020	8	6	easy	easy	JJ
hansen-can-2020	8	7	as	as	IN
hansen-can-2020	8	8	that	that	DT
hansen-can-2020	8	9	,	,	,
hansen-can-2020	8	10	and	and	CC
hansen-can-2020	8	11	numerous	numerous	JJ
hansen-can-2020	8	12	requests	request	NNS
hansen-can-2020	8	13	for	for	IN
hansen-can-2020	8	14	data	datum	NNS
hansen-can-2020	8	15	went	go	VBD
hansen-can-2020	8	16	without	without	IN
hansen-can-2020	8	17	response	response	NN
hansen-can-2020	8	18	.	.	.
hansen-can-2020	9	1	Even	even	RB
hansen-can-2020	9	2	after	after	IN
hansen-can-2020	9	3	I	-PRON-	PRP
hansen-can-2020	9	4	reached	reach	VBD
hansen-can-2020	9	5	out	out	RP
hansen-can-2020	9	6	to	to	IN
hansen-can-2020	9	7	personal	personal	JJ
hansen-can-2020	9	8	contacts	contact	NNS
hansen-can-2020	9	9	at	at	IN
hansen-can-2020	9	10	MathSciNet	MathSciNet	NNP
hansen-can-2020	9	11	,	,	,
hansen-can-2020	9	12	we	-PRON-	PRP
hansen-can-2020	9	13	came	come	VBD
hansen-can-2020	9	14	to	to	TO
hansen-can-2020	9	15	understand	understand	VB
hansen-can-2020	9	16	we	-PRON-	PRP
hansen-can-2020	9	17	would	would	MD
hansen-can-2020	9	18	not	not	RB
hansen-can-2020	9	19	be	be	VB
hansen-can-2020	9	20	getting	get	VBG
hansen-can-2020	9	21	the	the	DT
hansen-can-2020	9	22	MSC	MSC	NNP
hansen-can-2020	9	23	data	datum	NNS
hansen-can-2020	9	24	the	the	DT
hansen-can-2020	9	25	entire	entire	JJ
hansen-can-2020	9	26	project	project	NN
hansen-can-2020	9	27	relied	rely	VBN
hansen-can-2020	9	28	upon	upon	IN
hansen-can-2020	9	29	.	.	.
hansen-can-2020	10	1	Not	not	RB
hansen-can-2020	10	2	that	that	IN
hansen-can-2020	10	3	we	-PRON-	PRP
hansen-can-2020	10	4	were	be	VBD
hansen-can-2020	10	5	going	go	VBG
hansen-can-2020	10	6	to	to	TO
hansen-can-2020	10	7	let	let	VB
hansen-can-2020	10	8	a	a	DT
hansen-can-2020	10	9	little	little	JJ
hansen-can-2020	10	10	setback	setback	NN
hansen-can-2020	10	11	like	like	UH
hansen-can-2020	10	12	not	not	RB
hansen-can-2020	10	13	having	have	VBG
hansen-can-2020	10	14	the	the	DT
hansen-can-2020	10	15	necessary	necessary	JJ
hansen-can-2020	10	16	data	datum	NNS
hansen-can-2020	10	17	stop	stop	VB
hansen-can-2020	10	18	us	-PRON-	PRP
hansen-can-2020	10	19	.	.	.
hansen-can-2020	11	1	Comment	comment	NN
hansen-can-2020	11	2	by	by	IN
hansen-can-2020	11	3	Mark	Mark	NNP
hansen-can-2020	11	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	11	5	:	:	:
hansen-can-2020	11	6	Is	be	VBZ
hansen-can-2020	11	7	this	this	DT
hansen-can-2020	11	8	meant	mean	VBN
hansen-can-2020	11	9	to	to	TO
hansen-can-2020	11	10	be	be	VB
hansen-can-2020	11	11	MARC	MARC	NNP
hansen-can-2020	11	12	subfields	subfield	NNS
hansen-can-2020	11	13	or	or	CC
hansen-can-2020	11	14	professional	professional	JJ
hansen-can-2020	11	15	subfields	subfield	NNS
hansen-can-2020	11	16	?	?	.
hansen-can-2020	12	1	It	-PRON-	PRP
hansen-can-2020	12	2	might	may	MD
hansen-can-2020	12	3	be	be	VB
hansen-can-2020	12	4	good	good	JJ
hansen-can-2020	12	5	to	to	TO
hansen-can-2020	12	6	disambiguate	disambiguate	VB
hansen-can-2020	12	7	.	.	.
hansen-can-2020	13	1	[	[	-LRB-
hansen-can-2020	13	2	0	0	CD
hansen-can-2020	13	3	:	:	:
hansen-can-2020	13	4	https://mathscinet.ams.org/	https://mathscinet.ams.org/	NNP
hansen-can-2020	13	5	]	]	-RRB-
hansen-can-2020	13	6	After	after	RB
hansen-can-2020	13	7	all	all	RB
hansen-can-2020	13	8	,	,	,
hansen-can-2020	13	9	this	this	DT
hansen-can-2020	13	10	was	be	VBD
hansen-can-2020	13	11	early	early	JJ
hansen-can-2020	13	12	2018	2018	CD
hansen-can-2020	13	13	and	and	CC
hansen-can-2020	13	14	there	there	EX
hansen-can-2020	13	15	had	have	VBD
hansen-can-2020	13	16	already	already	RB
hansen-can-2020	13	17	been	be	VBN
hansen-can-2020	13	18	years	year	NNS
hansen-can-2020	13	19	of	of	IN
hansen-can-2020	13	20	stories	story	NNS
hansen-can-2020	13	21	about	about	IN
hansen-can-2020	13	22	how	how	WRB
hansen-can-2020	13	23	artificial	artificial	JJ
hansen-can-2020	13	24	intelligence	intelligence	NN
hansen-can-2020	13	25	,	,	,
hansen-can-2020	13	26	machine	machine	NN
hansen-can-2020	13	27	learning	learning	NN
hansen-can-2020	13	28	in	in	IN
hansen-can-2020	13	29	particular	particular	JJ
hansen-can-2020	13	30	,	,	,
hansen-can-2020	13	31	was	be	VBD
hansen-can-2020	13	32	going	go	VBG
hansen-can-2020	13	33	to	to	TO
hansen-can-2020	13	34	revolutionize	revolutionize	VB
hansen-can-2020	13	35	every	every	DT
hansen-can-2020	13	36	aspect	aspect	NN
hansen-can-2020	13	37	of	of	IN
hansen-can-2020	13	38	our	-PRON-	PRP$
hansen-can-2020	13	39	world	world	NN
hansen-can-2020	13	40	(	(	-LRB-
hansen-can-2020	13	41	Kelly	Kelly	NNP
hansen-can-2020	13	42	2014	2014	CD
hansen-can-2020	13	43	;	;	:
hansen-can-2020	13	44	Clark	Clark	NNP
hansen-can-2020	13	45	2015	2015	CD
hansen-can-2020	13	46	;	;	:
hansen-can-2020	13	47	Parloff	Parloff	NNP
hansen-can-2020	13	48	2016	2016	CD
hansen-can-2020	13	49	;	;	:
hansen-can-2020	13	50	Sangwani	Sangwani	NNP
hansen-can-2020	13	51	2017	2017	CD
hansen-can-2020	13	52	;	;	:
hansen-can-2020	13	53	Tank	tank	NN
hansen-can-2020	13	54	2017	2017	CD
hansen-can-2020	13	55	)	)	-RRB-
hansen-can-2020	13	56	.	.	.
hansen-can-2020	14	1	All	all	PDT
hansen-can-2020	14	2	the	the	DT
hansen-can-2020	14	3	coverage	coverage	NN
hansen-can-2020	14	4	made	make	VBD
hansen-can-2020	14	5	it	-PRON-	PRP
hansen-can-2020	14	6	seem	seem	VB
hansen-can-2020	14	7	like	like	IN
hansen-can-2020	14	8	AI	AI	NNP
hansen-can-2020	14	9	was	be	VBD
hansen-can-2020	14	10	not	not	RB
hansen-can-2020	14	11	only	only	RB
hansen-can-2020	14	12	a	a	DT
hansen-can-2020	14	13	tool	tool	NN
hansen-can-2020	14	14	with	with	IN
hansen-can-2020	14	15	as	as	RB
hansen-can-2020	14	16	many	many	JJ
hansen-can-2020	14	17	applications	application	NNS
hansen-can-2020	14	18	as	as	IN
hansen-can-2020	14	19	a	a	DT
hansen-can-2020	14	20	hammer	hammer	NN
hansen-can-2020	14	21	,	,	,
hansen-can-2020	14	22	but	but	CC
hansen-can-2020	14	23	that	that	IN
hansen-can-2020	14	24	it	-PRON-	PRP
hansen-can-2020	14	25	also	also	RB
hansen-can-2020	14	26	magically	magically	RB
hansen-can-2020	14	27	turned	turn	VBD
hansen-can-2020	14	28	all	all	DT
hansen-can-2020	14	29	problems	problem	NNS
hansen-can-2020	14	30	into	into	IN
hansen-can-2020	14	31	nails	nail	NNS
hansen-can-2020	14	32	.	.	.
hansen-can-2020	15	1	While	while	IN
hansen-can-2020	15	2	none	none	NN
hansen-can-2020	15	3	of	of	IN
hansen-can-2020	15	4	us	-PRON-	PRP
hansen-can-2020	15	5	were	be	VBD
hansen-can-2020	15	6	AI	AI	NNP
hansen-can-2020	15	7	experts	expert	NNS
hansen-can-2020	15	8	,	,	,
hansen-can-2020	15	9	we	-PRON-	PRP
hansen-can-2020	15	10	knew	know	VBD
hansen-can-2020	15	11	that	that	DT
hansen-can-2020	15	12	machine	machine	NN
hansen-can-2020	15	13	learning	learning	NN
hansen-can-2020	15	14	was	be	VBD
hansen-can-2020	15	15	supposed	suppose	VBN
hansen-can-2020	15	16	to	to	TO
hansen-can-2020	15	17	be	be	VB
hansen-can-2020	15	18	good	good	JJ
hansen-can-2020	15	19	at	at	IN
hansen-can-2020	15	20	classification	classification	NN
hansen-can-2020	15	21	and	and	CC
hansen-can-2020	15	22	categorization	categorization	NN
hansen-can-2020	15	23	.	.	.
hansen-can-2020	16	1	The	the	DT
hansen-can-2020	16	2	promise	promise	NN
hansen-can-2020	16	3	seemed	seem	VBD
hansen-can-2020	16	4	to	to	TO
hansen-can-2020	16	5	be	be	VB
hansen-can-2020	16	6	that	that	IN
hansen-can-2020	16	7	if	if	IN
hansen-can-2020	16	8	you	-PRON-	PRP
hansen-can-2020	16	9	had	have	VBD
hansen-can-2020	16	10	stacks	stack	NNS
hansen-can-2020	16	11	of	of	IN
hansen-can-2020	16	12	data	datum	NNS
hansen-can-2020	16	13	,	,	,
hansen-can-2020	16	14	a	a	DT
hansen-can-2020	16	15	machine	machine	NN
hansen-can-2020	16	16	learning	learn	VBG
hansen-can-2020	16	17	algorithm	algorithm	NN
hansen-can-2020	16	18	could	could	MD
hansen-can-2020	16	19	dive	dive	VB
hansen-can-2020	16	20	in	in	RP
hansen-can-2020	16	21	,	,	,
hansen-can-2020	16	22	find	find	VB
hansen-can-2020	16	23	the	the	DT
hansen-can-2020	16	24	needles	needle	NNS
hansen-can-2020	16	25	,	,	,
hansen-can-2020	16	26	and	and	CC
hansen-can-2020	16	27	arrange	arrange	VB
hansen-can-2020	16	28	them	-PRON-	PRP
hansen-can-2020	16	29	into	into	IN
hansen-can-2020	16	30	neatly	neatly	RB
hansen-can-2020	16	31	divided	divide	VBN
hansen-can-2020	16	32	piles	pile	NNS
hansen-can-2020	16	33	of	of	IN
hansen-can-2020	16	34	similar	similar	JJ
hansen-can-2020	16	35	sharpness	sharpness	NN
hansen-can-2020	16	36	and	and	CC
hansen-can-2020	16	37	length	length	NN
hansen-can-2020	16	38	.	.	.
hansen-can-2020	17	1	Not	not	RB
hansen-can-2020	17	2	only	only	RB
hansen-can-2020	17	3	that	that	DT
hansen-can-2020	17	4	,	,	,
hansen-can-2020	17	5	but	but	CC
hansen-can-2020	17	6	there	there	EX
hansen-can-2020	17	7	were	be	VBD
hansen-can-2020	17	8	pre	pre	VBN
hansen-can-2020	17	9	-	-	JJ
hansen-can-2020	17	10	built	build	VBN
hansen-can-2020	17	11	tools	tool	NNS
hansen-can-2020	17	12	that	that	WDT
hansen-can-2020	17	13	made	make	VBD
hansen-can-2020	17	14	it	-PRON-	PRP
hansen-can-2020	17	15	so	so	RB
hansen-can-2020	17	16	almost	almost	RB
hansen-can-2020	17	17	anyone	anyone	NN
hansen-can-2020	17	18	could	could	MD
hansen-can-2020	17	19	do	do	VB
hansen-can-2020	17	20	it	-PRON-	PRP
hansen-can-2020	17	21	.	.	.
hansen-can-2020	18	1	For	for	IN
hansen-can-2020	18	2	a	a	DT
hansen-can-2020	18	3	group	group	NN
hansen-can-2020	18	4	of	of	IN
hansen-can-2020	18	5	people	people	NNS
hansen-can-2020	18	6	whose	whose	WP$
hansen-can-2020	18	7	project	project	NN
hansen-can-2020	18	8	was	be	VBD
hansen-can-2020	18	9	on	on	IN
hansen-can-2020	18	10	life	life	NN
hansen-can-2020	18	11	support	support	NN
hansen-can-2020	18	12	because	because	IN
hansen-can-2020	18	13	we	-PRON-	PRP
hansen-can-2020	18	14	could	could	MD
hansen-can-2020	18	15	not	not	RB
hansen-can-2020	18	16	get	get	VB
hansen-can-2020	18	17	the	the	DT
hansen-can-2020	18	18	categorization	categorization	NN
hansen-can-2020	18	19	data	datum	NNS
hansen-can-2020	18	20	we	-PRON-	PRP
hansen-can-2020	18	21	needed	need	VBD
hansen-can-2020	18	22	,	,	,
hansen-can-2020	18	23	machine	machine	NN
hansen-can-2020	18	24	learning	learning	NN
hansen-can-2020	18	25	began	begin	VBD
hansen-can-2020	18	26	to	to	TO
hansen-can-2020	18	27	look	look	VB
hansen-can-2020	18	28	like	like	IN
hansen-can-2020	18	29	our	-PRON-	PRP$
hansen-can-2020	18	30	only	only	JJ
hansen-can-2020	18	31	potential	potential	JJ
hansen-can-2020	18	32	savior	savior	NN
hansen-can-2020	18	33	.	.	.
hansen-can-2020	19	1	So	so	RB
hansen-can-2020	19	2	,	,	,
hansen-can-2020	19	3	machine	machine	NN
hansen-can-2020	19	4	learning	learning	NN
hansen-can-2020	19	5	is	be	VBZ
hansen-can-2020	19	6	what	what	WP
hansen-can-2020	19	7	we	-PRON-	PRP
hansen-can-2020	19	8	did	do	VBD
hansen-can-2020	19	9	.	.	.
hansen-can-2020	20	1	Comment	comment	NN
hansen-can-2020	20	2	by	by	IN
hansen-can-2020	20	3	Mark	Mark	NNP
hansen-can-2020	20	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	20	5	:	:	:
hansen-can-2020	20	6	Maybe	maybe	RB
hansen-can-2020	20	7	pursued	pursue	VBN
hansen-can-2020	20	8	?	?	.
hansen-can-2020	21	1	Is	be	VBZ
hansen-can-2020	21	2	machine	machine	NN
hansen-can-2020	21	3	learning	learn	VBG
hansen-can-2020	21	4	an	an	DT
hansen-can-2020	21	5	action	action	NN
hansen-can-2020	21	6	or	or	CC
hansen-can-2020	21	7	a	a	DT
hansen-can-2020	21	8	technique	technique	NN
hansen-can-2020	21	9	?	?	.
hansen-can-2020	22	1	Comment	comment	NN
hansen-can-2020	22	2	by	by	IN
hansen-can-2020	22	3	Daniel	Daniel	NNP
hansen-can-2020	22	4	Johnson	Johnson	NNP
hansen-can-2020	22	5	:	:	:
hansen-can-2020	22	6	"	"	``
hansen-can-2020	22	7	used	use	VBN
hansen-can-2020	22	8	"	"	''
hansen-can-2020	22	9	would	would	MD
hansen-can-2020	22	10	work	work	VB
hansen-can-2020	22	11	too	too	RB
hansen-can-2020	22	12	.	.	.
hansen-can-2020	23	1	I	-PRON-	PRP
hansen-can-2020	23	2	will	will	MD
hansen-can-2020	23	3	not	not	RB
hansen-can-2020	23	4	go	go	VB
hansen-can-2020	23	5	too	too	RB
hansen-can-2020	23	6	deep	deep	RB
hansen-can-2020	23	7	into	into	IN
hansen-can-2020	23	8	the	the	DT
hansen-can-2020	23	9	actual	actual	JJ
hansen-can-2020	23	10	process	process	NN
hansen-can-2020	23	11	,	,	,
hansen-can-2020	23	12	but	but	CC
hansen-can-2020	23	13	I	-PRON-	PRP
hansen-can-2020	23	14	will	will	MD
hansen-can-2020	23	15	give	give	VB
hansen-can-2020	23	16	a	a	DT
hansen-can-2020	23	17	brief	brief	JJ
hansen-can-2020	23	18	outline	outline	NN
hansen-can-2020	23	19	of	of	IN
hansen-can-2020	23	20	the	the	DT
hansen-can-2020	23	21	techniques	technique	NNS
hansen-can-2020	23	22	we	-PRON-	PRP
hansen-can-2020	23	23	employed	employ	VBD
hansen-can-2020	23	24	.	.	.
hansen-can-2020	24	1	Machine	machine	NN
hansen-can-2020	24	2	-	-	HYPH
hansen-can-2020	24	3	learning	learning	NN
hansen-can-2020	24	4	-	-	HYPH
hansen-can-2020	24	5	based	base	VBN
hansen-can-2020	24	6	categorization	categorization	NN
hansen-can-2020	24	7	needs	need	VBZ
hansen-can-2020	24	8	data	datum	NNS
hansen-can-2020	24	9	to	to	TO
hansen-can-2020	24	10	classify	classify	VB
hansen-can-2020	24	11	,	,	,
hansen-can-2020	24	12	which	which	WDT
hansen-can-2020	24	13	in	in	IN
hansen-can-2020	24	14	our	-PRON-	PRP$
hansen-can-2020	24	15	case	case	NN
hansen-can-2020	24	16	were	be	VBD
hansen-can-2020	24	17	mathematics	mathematic	NNS
hansen-can-2020	24	18	publications	publication	NNS
hansen-can-2020	24	19	.	.	.
hansen-can-2020	25	1	While	while	IN
hansen-can-2020	25	2	this	this	DT
hansen-can-2020	25	3	can	can	MD
hansen-can-2020	25	4	be	be	VB
hansen-can-2020	25	5	done	do	VBN
hansen-can-2020	25	6	with	with	IN
hansen-can-2020	25	7	titles	title	NNS
hansen-can-2020	25	8	and	and	CC
hansen-can-2020	25	9	abstracts	abstract	NNS
hansen-can-2020	25	10	we	-PRON-	PRP
hansen-can-2020	25	11	wanted	want	VBD
hansen-can-2020	25	12	to	to	TO
hansen-can-2020	25	13	provide	provide	VB
hansen-can-2020	25	14	the	the	DT
hansen-can-2020	25	15	machine	machine	NN
hansen-can-2020	25	16	with	with	IN
hansen-can-2020	25	17	as	as	RB
hansen-can-2020	25	18	much	much	JJ
hansen-can-2020	25	19	data	datum	NNS
hansen-can-2020	25	20	as	as	IN
hansen-can-2020	25	21	we	-PRON-	PRP
hansen-can-2020	25	22	could	could	MD
hansen-can-2020	25	23	,	,	,
hansen-can-2020	25	24	so	so	RB
hansen-can-2020	25	25	we	-PRON-	PRP
hansen-can-2020	25	26	decided	decide	VBD
hansen-can-2020	25	27	to	to	TO
hansen-can-2020	25	28	work	work	VB
hansen-can-2020	25	29	with	with	IN
hansen-can-2020	25	30	full	full	JJ
hansen-can-2020	25	31	-	-	HYPH
hansen-can-2020	25	32	text	text	NN
hansen-can-2020	25	33	articles	article	NNS
hansen-can-2020	25	34	.	.	.
hansen-can-2020	26	1	Since	since	IN
hansen-can-2020	26	2	we	-PRON-	PRP
hansen-can-2020	26	3	were	be	VBD
hansen-can-2020	26	4	at	at	IN
hansen-can-2020	26	5	the	the	DT
hansen-can-2020	26	6	University	University	NNP
hansen-can-2020	26	7	of	of	IN
hansen-can-2020	26	8	Wisconsin	Wisconsin	NNP
hansen-can-2020	26	9	at	at	IN
hansen-can-2020	26	10	the	the	DT
hansen-can-2020	26	11	time	time	NN
hansen-can-2020	26	12	,	,	,
hansen-can-2020	26	13	we	-PRON-	PRP
hansen-can-2020	26	14	were	be	VBD
hansen-can-2020	26	15	able	able	JJ
hansen-can-2020	26	16	to	to	TO
hansen-can-2020	26	17	connect	connect	VB
hansen-can-2020	26	18	with	with	IN
hansen-can-2020	26	19	the	the	DT
hansen-can-2020	26	20	team	team	NN
hansen-can-2020	26	21	behind	behind	IN
hansen-can-2020	26	22	GeoDeepDive[footnoteRef:1	GeoDeepDive[footnoteRef:1	NNS
hansen-can-2020	26	23	]	]	-RRB-
hansen-can-2020	26	24	who	who	WP
hansen-can-2020	26	25	have	have	VBP
hansen-can-2020	26	26	agreements	agreement	NNS
hansen-can-2020	26	27	with	with	IN
hansen-can-2020	26	28	many	many	JJ
hansen-can-2020	26	29	publishers	publisher	NNS
hansen-can-2020	26	30	to	to	TO
hansen-can-2020	26	31	provide	provide	VB
hansen-can-2020	26	32	the	the	DT
hansen-can-2020	26	33	full	full	JJ
hansen-can-2020	26	34	text	text	NN
hansen-can-2020	26	35	of	of	IN
hansen-can-2020	26	36	articles	article	NNS
hansen-can-2020	26	37	for	for	IN
hansen-can-2020	26	38	text	text	NN
hansen-can-2020	26	39	and	and	CC
hansen-can-2020	26	40	data	datum	NNS
hansen-can-2020	26	41	mining	mining	NN
hansen-can-2020	26	42	research	research	NN
hansen-can-2020	26	43	(	(	-LRB-
hansen-can-2020	26	44	“	"	``
hansen-can-2020	26	45	GeoDeepDive	geodeepdive	NN
hansen-can-2020	26	46	:	:	:
hansen-can-2020	26	47	Project	Project	NNP
hansen-can-2020	26	48	Overview	Overview	NNP
hansen-can-2020	26	49	”	"	''
hansen-can-2020	26	50	n.d	n.d	NNP
hansen-can-2020	26	51	.	.	NNP
hansen-can-2020	26	52	)	)	-RRB-
hansen-can-2020	26	53	.	.	.
hansen-can-2020	27	1	GeoDeepDive	geodeepdive	NN
hansen-can-2020	27	2	provided	provide	VBD
hansen-can-2020	27	3	us	-PRON-	PRP
hansen-can-2020	27	4	with	with	IN
hansen-can-2020	27	5	the	the	DT
hansen-can-2020	27	6	full	full	JJ
hansen-can-2020	27	7	text	text	NN
hansen-can-2020	27	8	of	of	IN
hansen-can-2020	27	9	22,397	22,397	CD
hansen-can-2020	27	10	mathematics	mathematic	NNS
hansen-can-2020	27	11	articles	article	NNS
hansen-can-2020	27	12	which	which	WDT
hansen-can-2020	27	13	we	-PRON-	PRP
hansen-can-2020	27	14	used	use	VBD
hansen-can-2020	27	15	as	as	IN
hansen-can-2020	27	16	our	-PRON-	PRP$
hansen-can-2020	27	17	corpus	corpus	NN
hansen-can-2020	27	18	.	.	.
hansen-can-2020	28	1	In	in	IN
hansen-can-2020	28	2	order	order	NN
hansen-can-2020	28	3	to	to	TO
hansen-can-2020	28	4	classify	classify	VB
hansen-can-2020	28	5	these	these	DT
hansen-can-2020	28	6	articles	article	NNS
hansen-can-2020	28	7	,	,	,
hansen-can-2020	28	8	which	which	WDT
hansen-can-2020	28	9	were	be	VBD
hansen-can-2020	28	10	already	already	RB
hansen-can-2020	28	11	pre	pre	VBN
hansen-can-2020	28	12	-	-	VBN
hansen-can-2020	28	13	processed	process	VBN
hansen-can-2020	28	14	by	by	IN
hansen-can-2020	28	15	GeoDeepDive	GeoDeepDive	NNP
hansen-can-2020	28	16	with	with	IN
hansen-can-2020	28	17	CoreNLP,[footnoteRef:2	CoreNLP,[footnoteRef:2	NNPS
hansen-can-2020	28	18	]	]	-RRB-
hansen-can-2020	28	19	we	-PRON-	PRP
hansen-can-2020	28	20	first	first	RB
hansen-can-2020	28	21	used	use	VBD
hansen-can-2020	28	22	the	the	DT
hansen-can-2020	28	23	Python	Python	NNP
hansen-can-2020	28	24	package	package	NN
hansen-can-2020	28	25	Gensim[footnoteRef:3	Gensim[footnoteRef:3	NNP
hansen-can-2020	28	26	]	]	-RRB-
hansen-can-2020	28	27	to	to	TO
hansen-can-2020	28	28	process	process	VB
hansen-can-2020	28	29	the	the	DT
hansen-can-2020	28	30	articles	article	NNS
hansen-can-2020	28	31	into	into	IN
hansen-can-2020	28	32	a	a	DT
hansen-can-2020	28	33	Python	Python	NNP
hansen-can-2020	28	34	-	-	HYPH
hansen-can-2020	28	35	friendly	friendly	JJ
hansen-can-2020	28	36	format	format	NN
hansen-can-2020	28	37	and	and	CC
hansen-can-2020	28	38	to	to	TO
hansen-can-2020	28	39	remove	remove	VB
hansen-can-2020	28	40	stopwords	stopword	NNS
hansen-can-2020	28	41	.	.	.
hansen-can-2020	29	1	Then	then	RB
hansen-can-2020	29	2	we	-PRON-	PRP
hansen-can-2020	29	3	randomly	randomly	RB
hansen-can-2020	29	4	sampled	sample	VBD
hansen-can-2020	29	5	⅓	⅓	NNS
hansen-can-2020	29	6	of	of	IN
hansen-can-2020	29	7	the	the	DT
hansen-can-2020	29	8	corpus	corpus	NNP
hansen-can-2020	29	9	to	to	TO
hansen-can-2020	29	10	create	create	VB
hansen-can-2020	29	11	a	a	DT
hansen-can-2020	29	12	topic	topic	JJ
hansen-can-2020	29	13	model	model	NN
hansen-can-2020	29	14	using	use	VBG
hansen-can-2020	29	15	the	the	DT
hansen-can-2020	29	16	MALLET[footnoteRef:4	mallet[footnoteref:4	NN
hansen-can-2020	29	17	]	]	-RRB-
hansen-can-2020	29	18	topic	topic	NN
hansen-can-2020	29	19	modeling	modeling	NN
hansen-can-2020	29	20	tool	tool	NN
hansen-can-2020	29	21	.	.	.
hansen-can-2020	30	1	Finally	finally	RB
hansen-can-2020	30	2	,	,	,
hansen-can-2020	30	3	we	-PRON-	PRP
hansen-can-2020	30	4	applied	apply	VBD
hansen-can-2020	30	5	the	the	DT
hansen-can-2020	30	6	model	model	NN
hansen-can-2020	30	7	to	to	IN
hansen-can-2020	30	8	the	the	DT
hansen-can-2020	30	9	remaining	remain	VBG
hansen-can-2020	30	10	articles	article	NNS
hansen-can-2020	30	11	in	in	IN
hansen-can-2020	30	12	our	-PRON-	PRP$
hansen-can-2020	30	13	corpus	corpus	NN
hansen-can-2020	30	14	.	.	.
hansen-can-2020	31	1	We	-PRON-	PRP
hansen-can-2020	31	2	then	then	RB
hansen-can-2020	31	3	coded	code	VBD
hansen-can-2020	31	4	the	the	DT
hansen-can-2020	31	5	words	word	NNS
hansen-can-2020	31	6	within	within	IN
hansen-can-2020	31	7	the	the	DT
hansen-can-2020	31	8	generated	generate	VBN
hansen-can-2020	31	9	topics	topic	NNS
hansen-can-2020	31	10	to	to	IN
hansen-can-2020	31	11	subfields	subfield	NNS
hansen-can-2020	31	12	within	within	IN
hansen-can-2020	31	13	mathematics	mathematic	NNS
hansen-can-2020	31	14	and	and	CC
hansen-can-2020	31	15	used	use	VBD
hansen-can-2020	31	16	those	those	DT
hansen-can-2020	31	17	codes	code	NNS
hansen-can-2020	31	18	to	to	IN
hansen-can-2020	31	19	assign	assign	VB
hansen-can-2020	31	20	articles	article	NNS
hansen-can-2020	31	21	a	a	DT
hansen-can-2020	31	22	subfield	subfield	JJ
hansen-can-2020	31	23	category	category	NN
hansen-can-2020	31	24	.	.	.
hansen-can-2020	32	1	In	in	IN
hansen-can-2020	32	2	order	order	NN
hansen-can-2020	32	3	to	to	TO
hansen-can-2020	32	4	make	make	VB
hansen-can-2020	32	5	sure	sure	JJ
hansen-can-2020	32	6	our	-PRON-	PRP$
hansen-can-2020	32	7	results	result	NNS
hansen-can-2020	32	8	were	be	VBD
hansen-can-2020	32	9	not	not	RB
hansen-can-2020	32	10	just	just	RB
hansen-can-2020	32	11	a	a	DT
hansen-can-2020	32	12	one	one	CD
hansen-can-2020	32	13	-	-	HYPH
hansen-can-2020	32	14	off	off	NN
hansen-can-2020	32	15	,	,	,
hansen-can-2020	32	16	we	-PRON-	PRP
hansen-can-2020	32	17	repeated	repeat	VBD
hansen-can-2020	32	18	this	this	DT
hansen-can-2020	32	19	process	process	NN
hansen-can-2020	32	20	multiple	multiple	JJ
hansen-can-2020	32	21	times	time	NNS
hansen-can-2020	32	22	and	and	CC
hansen-can-2020	32	23	checked	check	VBD
hansen-can-2020	32	24	for	for	IN
hansen-can-2020	32	25	variance	variance	NN
hansen-can-2020	32	26	in	in	IN
hansen-can-2020	32	27	the	the	DT
hansen-can-2020	32	28	results	result	NNS
hansen-can-2020	32	29	.	.	.
hansen-can-2020	33	1	There	there	EX
hansen-can-2020	33	2	was	be	VBD
hansen-can-2020	33	3	none	none	NN
hansen-can-2020	33	4	,	,	,
hansen-can-2020	33	5	the	the	DT
hansen-can-2020	33	6	results	result	NNS
hansen-can-2020	33	7	were	be	VBD
hansen-can-2020	33	8	uniformly	uniformly	RB
hansen-can-2020	33	9	poor	poor	JJ
hansen-can-2020	33	10	.	.	.
hansen-can-2020	34	1	[	[	-LRB-
hansen-can-2020	34	2	1	1	CD
hansen-can-2020	34	3	:	:	:
hansen-can-2020	34	4	https://geodeepdive.org/	https://geodeepdive.org/	LS
hansen-can-2020	34	5	]	]	-RRB-
hansen-can-2020	34	6	[	[	-LRB-
hansen-can-2020	34	7	2	2	CD
hansen-can-2020	34	8	:	:	:
hansen-can-2020	34	9	https://stanfordnlp.github.io/CoreNLP/	https://stanfordnlp.github.io/CoreNLP/	NNS
hansen-can-2020	34	10	]	]	-RRB-
hansen-can-2020	34	11	[	[	-LRB-
hansen-can-2020	34	12	3	3	CD
hansen-can-2020	34	13	:	:	:
hansen-can-2020	34	14	https://radimrehurek.com/gensim/	https://radimrehurek.com/gensim/	NN
hansen-can-2020	34	15	]	]	-RRB-
hansen-can-2020	34	16	[	[	-LRB-
hansen-can-2020	34	17	4	4	CD
hansen-can-2020	34	18	:	:	SYM
hansen-can-2020	34	19	http://mallet.cs.umass.edu/topics.php	http://mallet.cs.umass.edu/topics.php	LS
hansen-can-2020	34	20	]	]	-RRB-
hansen-can-2020	34	21	That	that	DT
hansen-can-2020	34	22	might	may	MD
hansen-can-2020	34	23	not	not	RB
hansen-can-2020	34	24	be	be	VB
hansen-can-2020	34	25	entirely	entirely	RB
hansen-can-2020	34	26	fair	fair	JJ
hansen-can-2020	34	27	.	.	.
hansen-can-2020	35	1	There	there	EX
hansen-can-2020	35	2	were	be	VBD
hansen-can-2020	35	3	interesting	interesting	JJ
hansen-can-2020	35	4	aspects	aspect	NNS
hansen-can-2020	35	5	to	to	IN
hansen-can-2020	35	6	the	the	DT
hansen-can-2020	35	7	results	result	NNS
hansen-can-2020	35	8	of	of	IN
hansen-can-2020	35	9	the	the	DT
hansen-can-2020	35	10	topic	topic	NN
hansen-can-2020	35	11	modeling	modeling	NN
hansen-can-2020	35	12	,	,	,
hansen-can-2020	35	13	but	but	CC
hansen-can-2020	35	14	when	when	WRB
hansen-can-2020	35	15	it	-PRON-	PRP
hansen-can-2020	35	16	came	come	VBD
hansen-can-2020	35	17	to	to	IN
hansen-can-2020	35	18	categorization	categorization	NN
hansen-can-2020	35	19	they	-PRON-	PRP
hansen-can-2020	35	20	were	be	VBD
hansen-can-2020	35	21	useless	useless	JJ
hansen-can-2020	35	22	.	.	.
hansen-can-2020	36	1	Of	of	IN
hansen-can-2020	36	2	the	the	DT
hansen-can-2020	36	3	subfield	subfield	NN
hansen-can-2020	36	4	codes	code	NNS
hansen-can-2020	36	5	assigned	assign	VBN
hansen-can-2020	36	6	to	to	IN
hansen-can-2020	36	7	articles	article	NNS
hansen-can-2020	36	8	,	,	,
hansen-can-2020	36	9	only	only	RB
hansen-can-2020	36	10	two	two	CD
hansen-can-2020	36	11	were	be	VBD
hansen-can-2020	36	12	ever	ever	RB
hansen-can-2020	36	13	the	the	DT
hansen-can-2020	36	14	dominant	dominant	JJ
hansen-can-2020	36	15	result	result	NN
hansen-can-2020	36	16	for	for	IN
hansen-can-2020	36	17	any	any	DT
hansen-can-2020	36	18	given	give	VBN
hansen-can-2020	36	19	article	article	NN
hansen-can-2020	36	20	:	:	:
hansen-can-2020	36	21	Graph	Graph	NNP
hansen-can-2020	36	22	Theory	Theory	NNP
hansen-can-2020	36	23	and	and	CC
hansen-can-2020	36	24	Undefined	Undefined	NNP
hansen-can-2020	36	25	,	,	,
hansen-can-2020	36	26	which	which	WDT
hansen-can-2020	36	27	does	do	VBZ
hansen-can-2020	36	28	not	not	RB
hansen-can-2020	36	29	really	really	RB
hansen-can-2020	36	30	tell	tell	VB
hansen-can-2020	36	31	the	the	DT
hansen-can-2020	36	32	whole	whole	JJ
hansen-can-2020	36	33	story	story	NN
hansen-can-2020	36	34	as	as	IN
hansen-can-2020	36	35	Undefined	Undefined	NNP
hansen-can-2020	36	36	was	be	VBD
hansen-can-2020	36	37	the	the	DT
hansen-can-2020	36	38	run	run	NN
hansen-can-2020	36	39	-	-	HYPH
hansen-can-2020	36	40	away	away	RP
hansen-can-2020	36	41	winner	winner	NN
hansen-can-2020	36	42	in	in	IN
hansen-can-2020	36	43	the	the	DT
hansen-can-2020	36	44	article	article	NN
hansen-can-2020	36	45	classification	classification	NN
hansen-can-2020	36	46	race	race	NN
hansen-can-2020	36	47	with	with	IN
hansen-can-2020	36	48	more	more	JJR
hansen-can-2020	36	49	than	than	IN
hansen-can-2020	36	50	70	70	CD
hansen-can-2020	36	51	%	%	NN
hansen-can-2020	36	52	of	of	IN
hansen-can-2020	36	53	articles	article	NNS
hansen-can-2020	36	54	classified	classify	VBN
hansen-can-2020	36	55	as	as	IN
hansen-can-2020	36	56	Undefined	undefined	JJ
hansen-can-2020	36	57	in	in	IN
hansen-can-2020	36	58	each	each	DT
hansen-can-2020	36	59	run	run	NN
hansen-can-2020	36	60	,	,	,
hansen-can-2020	36	61	including	include	VBG
hansen-can-2020	36	62	one	one	CD
hansen-can-2020	36	63	for	for	IN
hansen-can-2020	36	64	which	which	WDT
hansen-can-2020	36	65	it	-PRON-	PRP
hansen-can-2020	36	66	hit	hit	VBD
hansen-can-2020	36	67	95	95	CD
hansen-can-2020	36	68	%	%	NN
hansen-can-2020	36	69	.	.	.
hansen-can-2020	37	1	The	the	DT
hansen-can-2020	37	2	topics	topic	NNS
hansen-can-2020	37	3	generated	generate	VBN
hansen-can-2020	37	4	by	by	IN
hansen-can-2020	37	5	MALLET	MALLET	NNP
hansen-can-2020	37	6	were	be	VBD
hansen-can-2020	37	7	often	often	RB
hansen-can-2020	37	8	plagued	plague	VBN
hansen-can-2020	37	9	by	by	IN
hansen-can-2020	37	10	gibberish	gibberish	NN
hansen-can-2020	37	11	caused	cause	VBN
hansen-can-2020	37	12	by	by	IN
hansen-can-2020	37	13	equations	equation	NNS
hansen-can-2020	37	14	in	in	IN
hansen-can-2020	37	15	the	the	DT
hansen-can-2020	37	16	mathematics	mathematic	NNS
hansen-can-2020	37	17	articles	article	NNS
hansen-can-2020	37	18	and	and	CC
hansen-can-2020	37	19	there	there	EX
hansen-can-2020	37	20	was	be	VBD
hansen-can-2020	37	21	at	at	RB
hansen-can-2020	37	22	least	least	RBS
hansen-can-2020	37	23	one	one	CD
hansen-can-2020	37	24	topic	topic	NN
hansen-can-2020	37	25	in	in	IN
hansen-can-2020	37	26	each	each	DT
hansen-can-2020	37	27	run	run	NN
hansen-can-2020	37	28	that	that	WDT
hansen-can-2020	37	29	was	be	VBD
hansen-can-2020	37	30	filled	fill	VBN
hansen-can-2020	37	31	with	with	IN
hansen-can-2020	37	32	the	the	DT
hansen-can-2020	37	33	names	name	NNS
hansen-can-2020	37	34	of	of	IN
hansen-can-2020	37	35	months	month	NNS
hansen-can-2020	37	36	and	and	CC
hansen-can-2020	37	37	locations	location	NNS
hansen-can-2020	37	38	.	.	.
hansen-can-2020	38	1	Add	add	VB
hansen-can-2020	38	2	how	how	WRB
hansen-can-2020	38	3	the	the	DT
hansen-can-2020	38	4	technical	technical	JJ
hansen-can-2020	38	5	language	language	NN
hansen-can-2020	38	6	of	of	IN
hansen-can-2020	38	7	mathematics	mathematic	NNS
hansen-can-2020	38	8	is	be	VBZ
hansen-can-2020	38	9	filled	fill	VBN
hansen-can-2020	38	10	with	with	IN
hansen-can-2020	38	11	words	word	NNS
hansen-can-2020	38	12	that	that	WDT
hansen-can-2020	38	13	have	have	VBP
hansen-can-2020	38	14	non	non	JJ
hansen-can-2020	38	15	-	-	JJ
hansen-can-2020	38	16	technical	technical	JJ
hansen-can-2020	38	17	definitions	definition	NNS
hansen-can-2020	38	18	(	(	-LRB-
hansen-can-2020	38	19	for	for	IN
hansen-can-2020	38	20	example	example	NN
hansen-can-2020	38	21	,	,	,
hansen-can-2020	38	22	map	map	NN
hansen-can-2020	38	23	or	or	CC
hansen-can-2020	38	24	space	space	NN
hansen-can-2020	38	25	)	)	-RRB-
hansen-can-2020	38	26	,	,	,
hansen-can-2020	38	27	or	or	CC
hansen-can-2020	38	28	words	word	NNS
hansen-can-2020	38	29	which	which	WDT
hansen-can-2020	38	30	have	have	VBP
hansen-can-2020	38	31	their	-PRON-	PRP$
hansen-can-2020	38	32	own	own	JJ
hansen-can-2020	38	33	subfield	subfield	NN
hansen-can-2020	38	34	-	-	HYPH
hansen-can-2020	38	35	specific	specific	JJ
hansen-can-2020	38	36	meanings	meaning	NNS
hansen-can-2020	38	37	(	(	-LRB-
hansen-can-2020	38	38	such	such	JJ
hansen-can-2020	38	39	as	as	IN
hansen-can-2020	38	40	homomorphism	homomorphism	NN
hansen-can-2020	38	41	or	or	CC
hansen-can-2020	38	42	degree	degree	NN
hansen-can-2020	38	43	)	)	-RRB-
hansen-can-2020	38	44	,	,	,
hansen-can-2020	38	45	both	both	DT
hansen-can-2020	38	46	of	of	IN
hansen-can-2020	38	47	which	which	WDT
hansen-can-2020	38	48	frustrate	frustrate	VBP
hansen-can-2020	38	49	attempts	attempt	NNS
hansen-can-2020	38	50	to	to	TO
hansen-can-2020	38	51	code	code	VB
hansen-can-2020	38	52	a	a	DT
hansen-can-2020	38	53	subfield	subfield	NN
hansen-can-2020	38	54	.	.	.
hansen-can-2020	39	1	These	these	DT
hansen-can-2020	39	2	issues	issue	NNS
hansen-can-2020	39	3	help	help	VBP
hansen-can-2020	39	4	make	make	VB
hansen-can-2020	39	5	it	-PRON-	PRP
hansen-can-2020	39	6	clear	clear	JJ
hansen-can-2020	39	7	why	why	WRB
hansen-can-2020	39	8	so	so	RB
hansen-can-2020	39	9	many	many	JJ
hansen-can-2020	39	10	articles	article	NNS
hansen-can-2020	39	11	ended	end	VBD
hansen-can-2020	39	12	up	up	RP
hansen-can-2020	39	13	as	as	IN
hansen-can-2020	39	14	“	"	``
hansen-can-2020	39	15	Undefined	undefined	JJ
hansen-can-2020	39	16	.	.	.
hansen-can-2020	39	17	”	"	''
hansen-can-2020	39	18	Even	even	RB
hansen-can-2020	39	19	for	for	IN
hansen-can-2020	39	20	the	the	DT
hansen-can-2020	39	21	one	one	CD
hansen-can-2020	39	22	subfield	subfield	NN
hansen-can-2020	39	23	which	which	WDT
hansen-can-2020	39	24	had	have	VBD
hansen-can-2020	39	25	a	a	DT
hansen-can-2020	39	26	unique	unique	JJ
hansen-can-2020	39	27	enough	enough	JJ
hansen-can-2020	39	28	vocabulary	vocabulary	NN
hansen-can-2020	39	29	for	for	IN
hansen-can-2020	39	30	our	-PRON-	PRP$
hansen-can-2020	39	31	topic	topic	JJ
hansen-can-2020	39	32	model	model	NN
hansen-can-2020	39	33	to	to	TO
hansen-can-2020	39	34	partially	partially	RB
hansen-can-2020	39	35	be	be	VB
hansen-can-2020	39	36	able	able	JJ
hansen-can-2020	39	37	to	to	TO
hansen-can-2020	39	38	identify	identify	VB
hansen-can-2020	39	39	,	,	,
hansen-can-2020	39	40	Graph	Graph	NNP
hansen-can-2020	39	41	Theory	Theory	NNP
hansen-can-2020	39	42	,	,	,
hansen-can-2020	39	43	the	the	DT
hansen-can-2020	39	44	results	result	NNS
hansen-can-2020	39	45	were	be	VBD
hansen-can-2020	39	46	marginally	marginally	RB
hansen-can-2020	39	47	positive	positive	JJ
hansen-can-2020	39	48	at	at	IN
hansen-can-2020	39	49	best	good	JJS
hansen-can-2020	39	50	.	.	.
hansen-can-2020	40	1	We	-PRON-	PRP
hansen-can-2020	40	2	were	be	VBD
hansen-can-2020	40	3	able	able	JJ
hansen-can-2020	40	4	to	to	TO
hansen-can-2020	40	5	obtain	obtain	VB
hansen-can-2020	40	6	Mathematical	Mathematical	NNP
hansen-can-2020	40	7	Subject	Subject	NNP
hansen-can-2020	40	8	Classification	Classification	NNP
hansen-can-2020	40	9	(	(	-LRB-
hansen-can-2020	40	10	MSC	MSC	NNP
hansen-can-2020	40	11	)	)	-RRB-
hansen-can-2020	40	12	values	value	VBZ
hansen-can-2020	40	13	for	for	IN
hansen-can-2020	40	14	around	around	IN
hansen-can-2020	40	15	10	10	CD
hansen-can-2020	40	16	%	%	NN
hansen-can-2020	40	17	of	of	IN
hansen-can-2020	40	18	our	-PRON-	PRP$
hansen-can-2020	40	19	corpus	corpus	NN
hansen-can-2020	40	20	.	.	.
hansen-can-2020	41	1	When	when	WRB
hansen-can-2020	41	2	we	-PRON-	PRP
hansen-can-2020	41	3	compared	compare	VBD
hansen-can-2020	41	4	the	the	DT
hansen-can-2020	41	5	articles	article	NNS
hansen-can-2020	41	6	we	-PRON-	PRP
hansen-can-2020	41	7	categorized	categorize	VBD
hansen-can-2020	41	8	as	as	IN
hansen-can-2020	41	9	Graph	Graph	NNP
hansen-can-2020	41	10	Theory	Theory	NNP
hansen-can-2020	41	11	to	to	IN
hansen-can-2020	41	12	the	the	DT
hansen-can-2020	41	13	articles	article	NNS
hansen-can-2020	41	14	which	which	WDT
hansen-can-2020	41	15	had	have	VBD
hansen-can-2020	41	16	been	be	VBN
hansen-can-2020	41	17	assigned	assign	VBN
hansen-can-2020	41	18	the	the	DT
hansen-can-2020	41	19	MSC	MSC	NNP
hansen-can-2020	41	20	value	value	NN
hansen-can-2020	41	21	for	for	IN
hansen-can-2020	41	22	Graph	Graph	NNP
hansen-can-2020	41	23	Theory	Theory	NNP
hansen-can-2020	41	24	(	(	-LRB-
hansen-can-2020	41	25	05Cxx	05Cxx	NNP
hansen-can-2020	41	26	)	)	-RRB-
hansen-can-2020	41	27	,	,	,
hansen-can-2020	41	28	we	-PRON-	PRP
hansen-can-2020	41	29	found	find	VBD
hansen-can-2020	41	30	we	-PRON-	PRP
hansen-can-2020	41	31	had	have	VBD
hansen-can-2020	41	32	a	a	DT
hansen-can-2020	41	33	textbook	textbook	NN
hansen-can-2020	41	34	recall	recall	NN
hansen-can-2020	41	35	-	-	HYPH
hansen-can-2020	41	36	versus	versus	IN
hansen-can-2020	41	37	-	-	HYPH
hansen-can-2020	41	38	precision	precision	NN
hansen-can-2020	41	39	problem	problem	NN
hansen-can-2020	41	40	.	.	.
hansen-can-2020	42	1	We	-PRON-	PRP
hansen-can-2020	42	2	could	could	MD
hansen-can-2020	42	3	either	either	RB
hansen-can-2020	42	4	correctly	correctly	RB
hansen-can-2020	42	5	categorize	categorize	VB
hansen-can-2020	42	6	nearly	nearly	RB
hansen-can-2020	42	7	all	all	DT
hansen-can-2020	42	8	of	of	IN
hansen-can-2020	42	9	the	the	DT
hansen-can-2020	42	10	Graph	Graph	NNP
hansen-can-2020	42	11	Theory	Theory	NNP
hansen-can-2020	42	12	articles	article	VBZ
hansen-can-2020	42	13	with	with	IN
hansen-can-2020	42	14	a	a	DT
hansen-can-2020	42	15	very	very	RB
hansen-can-2020	42	16	high	high	JJ
hansen-can-2020	42	17	rate	rate	NN
hansen-can-2020	42	18	of	of	IN
hansen-can-2020	42	19	false	false	JJ
hansen-can-2020	42	20	positives	positive	NNS
hansen-can-2020	42	21	(	(	-LRB-
hansen-can-2020	42	22	high	high	JJ
hansen-can-2020	42	23	recall	recall	NN
hansen-can-2020	42	24	and	and	CC
hansen-can-2020	42	25	low	low	JJ
hansen-can-2020	42	26	precision	precision	NN
hansen-can-2020	42	27	)	)	-RRB-
hansen-can-2020	42	28	or	or	CC
hansen-can-2020	42	29	we	-PRON-	PRP
hansen-can-2020	42	30	could	could	MD
hansen-can-2020	42	31	almost	almost	RB
hansen-can-2020	42	32	never	never	RB
hansen-can-2020	42	33	incorrectly	incorrectly	RB
hansen-can-2020	42	34	categorize	categorize	VB
hansen-can-2020	42	35	an	an	DT
hansen-can-2020	42	36	article	article	NN
hansen-can-2020	42	37	as	as	IN
hansen-can-2020	42	38	Graph	Graph	NNP
hansen-can-2020	42	39	Theory	Theory	NNP
hansen-can-2020	42	40	,	,	,
hansen-can-2020	42	41	but	but	CC
hansen-can-2020	42	42	miss	miss	VB
hansen-can-2020	42	43	over	over	IN
hansen-can-2020	42	44	30	30	CD
hansen-can-2020	42	45	%	%	NN
hansen-can-2020	42	46	that	that	IN
hansen-can-2020	42	47	we	-PRON-	PRP
hansen-can-2020	42	48	should	should	MD
hansen-can-2020	42	49	have	have	VB
hansen-can-2020	42	50	categorized	categorize	VBN
hansen-can-2020	42	51	as	as	IN
hansen-can-2020	42	52	Graph	Graph	NNP
hansen-can-2020	42	53	Theory	Theory	NNP
hansen-can-2020	42	54	(	(	-LRB-
hansen-can-2020	42	55	high	high	JJ
hansen-can-2020	42	56	precision	precision	NN
hansen-can-2020	42	57	and	and	CC
hansen-can-2020	42	58	low	low	JJ
hansen-can-2020	42	59	recall	recall	NN
hansen-can-2020	42	60	)	)	-RRB-
hansen-can-2020	42	61	.	.	.
hansen-can-2020	43	1	Needless	needless	JJ
hansen-can-2020	43	2	to	to	TO
hansen-can-2020	43	3	say	say	VB
hansen-can-2020	43	4	,	,	,
hansen-can-2020	43	5	we	-PRON-	PRP
hansen-can-2020	43	6	were	be	VBD
hansen-can-2020	43	7	not	not	RB
hansen-can-2020	43	8	able	able	JJ
hansen-can-2020	43	9	to	to	TO
hansen-can-2020	43	10	create	create	VB
hansen-can-2020	43	11	the	the	DT
hansen-can-2020	43	12	temporal	temporal	JJ
hansen-can-2020	43	13	subfield	subfield	NN
hansen-can-2020	43	14	network	network	NN
hansen-can-2020	43	15	I	-PRON-	PRP
hansen-can-2020	43	16	had	have	VBD
hansen-can-2020	43	17	imagined	imagine	VBN
hansen-can-2020	43	18	.	.	.
hansen-can-2020	44	1	While	while	IN
hansen-can-2020	44	2	we	-PRON-	PRP
hansen-can-2020	44	3	could	could	MD
hansen-can-2020	44	4	reasonably	reasonably	RB
hansen-can-2020	44	5	claim	claim	VB
hansen-can-2020	44	6	that	that	IN
hansen-can-2020	44	7	we	-PRON-	PRP
hansen-can-2020	44	8	learned	learn	VBD
hansen-can-2020	44	9	very	very	RB
hansen-can-2020	44	10	interesting	interesting	JJ
hansen-can-2020	44	11	things	thing	NNS
hansen-can-2020	44	12	about	about	IN
hansen-can-2020	44	13	the	the	DT
hansen-can-2020	44	14	language	language	NN
hansen-can-2020	44	15	of	of	IN
hansen-can-2020	44	16	mathematics	mathematic	NNS
hansen-can-2020	44	17	and	and	CC
hansen-can-2020	44	18	its	-PRON-	PRP$
hansen-can-2020	44	19	subfields	subfield	NNS
hansen-can-2020	44	20	,	,	,
hansen-can-2020	44	21	we	-PRON-	PRP
hansen-can-2020	44	22	could	could	MD
hansen-can-2020	44	23	not	not	RB
hansen-can-2020	44	24	claim	claim	VB
hansen-can-2020	44	25	we	-PRON-	PRP
hansen-can-2020	44	26	even	even	RB
hansen-can-2020	44	27	came	come	VBD
hansen-can-2020	44	28	close	close	RB
hansen-can-2020	44	29	to	to	IN
hansen-can-2020	44	30	automatically	automatically	RB
hansen-can-2020	44	31	categorizing	categorize	VBG
hansen-can-2020	44	32	mathematics	mathematic	NNS
hansen-can-2020	44	33	articles	article	NNS
hansen-can-2020	44	34	.	.	.
hansen-can-2020	45	1	When	when	WRB
hansen-can-2020	45	2	we	-PRON-	PRP
hansen-can-2020	45	3	had	have	VBD
hansen-can-2020	45	4	to	to	TO
hansen-can-2020	45	5	report	report	VB
hansen-can-2020	45	6	back	back	RP
hansen-can-2020	45	7	on	on	IN
hansen-can-2020	45	8	our	-PRON-	PRP$
hansen-can-2020	45	9	work	work	NN
hansen-can-2020	45	10	at	at	IN
hansen-can-2020	45	11	the	the	DT
hansen-can-2020	45	12	end	end	NN
hansen-can-2020	45	13	of	of	IN
hansen-can-2020	45	14	the	the	DT
hansen-can-2020	45	15	course	course	NN
hansen-can-2020	45	16	,	,	,
hansen-can-2020	45	17	our	-PRON-	PRP$
hansen-can-2020	45	18	main	main	JJ
hansen-can-2020	45	19	result	result	NN
hansen-can-2020	45	20	was	be	VBD
hansen-can-2020	45	21	that	that	IN
hansen-can-2020	45	22	basic	basic	JJ
hansen-can-2020	45	23	,	,	,
hansen-can-2020	45	24	off	off	IN
hansen-can-2020	45	25	-	-	HYPH
hansen-can-2020	45	26	the	the	DT
hansen-can-2020	45	27	-	-	HYPH
hansen-can-2020	45	28	shelf	shelf	NN
hansen-can-2020	45	29	topic	topic	NN
hansen-can-2020	45	30	modelling	modelling	NN
hansen-can-2020	45	31	does	do	VBZ
hansen-can-2020	45	32	not	not	RB
hansen-can-2020	45	33	work	work	VB
hansen-can-2020	45	34	well	well	RB
hansen-can-2020	45	35	when	when	WRB
hansen-can-2020	45	36	it	-PRON-	PRP
hansen-can-2020	45	37	comes	come	VBZ
hansen-can-2020	45	38	to	to	IN
hansen-can-2020	45	39	highly	highly	RB
hansen-can-2020	45	40	technical	technical	JJ
hansen-can-2020	45	41	articles	article	NNS
hansen-can-2020	45	42	from	from	IN
hansen-can-2020	45	43	subjects	subject	NNS
hansen-can-2020	45	44	like	like	IN
hansen-can-2020	45	45	mathematics	mathematic	NNS
hansen-can-2020	45	46	.	.	.
hansen-can-2020	46	1	It	-PRON-	PRP
hansen-can-2020	46	2	was	be	VBD
hansen-can-2020	46	3	also	also	RB
hansen-can-2020	46	4	a	a	DT
hansen-can-2020	46	5	welcome	welcome	JJ
hansen-can-2020	46	6	lesson	lesson	NN
hansen-can-2020	46	7	in	in	IN
hansen-can-2020	46	8	not	not	RB
hansen-can-2020	46	9	believing	believe	VBG
hansen-can-2020	46	10	the	the	DT
hansen-can-2020	46	11	hype	hype	NN
hansen-can-2020	46	12	of	of	IN
hansen-can-2020	46	13	machine	machine	NN
hansen-can-2020	46	14	learning	learning	NN
hansen-can-2020	46	15	,	,	,
hansen-can-2020	46	16	even	even	RB
hansen-can-2020	46	17	when	when	WRB
hansen-can-2020	46	18	a	a	DT
hansen-can-2020	46	19	project	project	NN
hansen-can-2020	46	20	looks	look	VBZ
hansen-can-2020	46	21	exactly	exactly	RB
hansen-can-2020	46	22	like	like	IN
hansen-can-2020	46	23	the	the	DT
hansen-can-2020	46	24	kind	kind	JJ
hansen-can-2020	46	25	machine	machine	NN
hansen-can-2020	46	26	learning	learning	NN
hansen-can-2020	46	27	was	be	VBD
hansen-can-2020	46	28	supposed	suppose	VBN
hansen-can-2020	46	29	to	to	TO
hansen-can-2020	46	30	excel	excel	VB
hansen-can-2020	46	31	at	at	IN
hansen-can-2020	46	32	solving	solve	VBG
hansen-can-2020	46	33	.	.	.
hansen-can-2020	47	1	While	while	IN
hansen-can-2020	47	2	we	-PRON-	PRP
hansen-can-2020	47	3	had	have	VBD
hansen-can-2020	47	4	a	a	DT
hansen-can-2020	47	5	hammer	hammer	NN
hansen-can-2020	47	6	and	and	CC
hansen-can-2020	47	7	our	-PRON-	PRP$
hansen-can-2020	47	8	problem	problem	NN
hansen-can-2020	47	9	looked	look	VBD
hansen-can-2020	47	10	like	like	IN
hansen-can-2020	47	11	a	a	DT
hansen-can-2020	47	12	nail	nail	NN
hansen-can-2020	47	13	,	,	,
hansen-can-2020	47	14	it	-PRON-	PRP
hansen-can-2020	47	15	seemed	seem	VBD
hansen-can-2020	47	16	that	that	IN
hansen-can-2020	47	17	the	the	DT
hansen-can-2020	47	18	former	former	JJ
hansen-can-2020	47	19	was	be	VBD
hansen-can-2020	47	20	a	a	DT
hansen-can-2020	47	21	ball	ball	NN
hansen-can-2020	47	22	peen	peen	NN
hansen-can-2020	47	23	and	and	CC
hansen-can-2020	47	24	the	the	DT
hansen-can-2020	47	25	latter	latter	JJ
hansen-can-2020	47	26	a	a	DT
hansen-can-2020	47	27	railroad	railroad	NN
hansen-can-2020	47	28	tie	tie	NN
hansen-can-2020	47	29	.	.	.
hansen-can-2020	48	1	In	in	IN
hansen-can-2020	48	2	the	the	DT
hansen-can-2020	48	3	end	end	NN
hansen-can-2020	48	4	,	,	,
hansen-can-2020	48	5	even	even	RB
hansen-can-2020	48	6	in	in	IN
hansen-can-2020	48	7	the	the	DT
hansen-can-2020	48	8	land	land	NN
hansen-can-2020	48	9	of	of	IN
hansen-can-2020	48	10	hammers	hammer	NNS
hansen-can-2020	48	11	and	and	CC
hansen-can-2020	48	12	nails	nail	NNS
hansen-can-2020	48	13	,	,	,
hansen-can-2020	48	14	the	the	DT
hansen-can-2020	48	15	tool	tool	NN
hansen-can-2020	48	16	has	have	VBZ
hansen-can-2020	48	17	to	to	TO
hansen-can-2020	48	18	match	match	VB
hansen-can-2020	48	19	the	the	DT
hansen-can-2020	48	20	task	task	NN
hansen-can-2020	48	21	.	.	.
hansen-can-2020	49	1	Though	though	IN
hansen-can-2020	49	2	we	-PRON-	PRP
hansen-can-2020	49	3	failed	fail	VBD
hansen-can-2020	49	4	to	to	TO
hansen-can-2020	49	5	accomplish	accomplish	VB
hansen-can-2020	49	6	automated	automate	VBN
hansen-can-2020	49	7	categorization	categorization	NN
hansen-can-2020	49	8	of	of	IN
hansen-can-2020	49	9	mathematics	mathematic	NNS
hansen-can-2020	49	10	,	,	,
hansen-can-2020	49	11	we	-PRON-	PRP
hansen-can-2020	49	12	were	be	VBD
hansen-can-2020	49	13	dilettantes	dilettante	NNS
hansen-can-2020	49	14	in	in	IN
hansen-can-2020	49	15	the	the	DT
hansen-can-2020	49	16	world	world	NN
hansen-can-2020	49	17	of	of	IN
hansen-can-2020	49	18	machine	machine	NN
hansen-can-2020	49	19	learning	learning	NN
hansen-can-2020	49	20	.	.	.
hansen-can-2020	50	1	I	-PRON-	PRP
hansen-can-2020	50	2	believe	believe	VBP
hansen-can-2020	50	3	our	-PRON-	PRP$
hansen-can-2020	50	4	project	project	NN
hansen-can-2020	50	5	is	be	VBZ
hansen-can-2020	50	6	a	a	DT
hansen-can-2020	50	7	good	good	JJ
hansen-can-2020	50	8	example	example	NN
hansen-can-2020	50	9	of	of	IN
hansen-can-2020	50	10	how	how	WRB
hansen-can-2020	50	11	machine	machine	NN
hansen-can-2020	50	12	learning	learning	NN
hansen-can-2020	50	13	is	be	VBZ
hansen-can-2020	50	14	still	still	RB
hansen-can-2020	50	15	a	a	DT
hansen-can-2020	50	16	long	long	JJ
hansen-can-2020	50	17	way	way	NN
hansen-can-2020	50	18	from	from	IN
hansen-can-2020	50	19	being	be	VBG
hansen-can-2020	50	20	the	the	DT
hansen-can-2020	50	21	magic	magic	JJ
hansen-can-2020	50	22	tool	tool	NN
hansen-can-2020	50	23	as	as	IN
hansen-can-2020	50	24	some	some	DT
hansen-can-2020	50	25	,	,	,
hansen-can-2020	50	26	though	though	IN
hansen-can-2020	50	27	not	not	RB
hansen-can-2020	50	28	all	all	DT
hansen-can-2020	50	29	(	(	-LRB-
hansen-can-2020	50	30	Rahimi	Rahimi	NNP
hansen-can-2020	50	31	and	and	CC
hansen-can-2020	50	32	Recht	Recht	NNP
hansen-can-2020	50	33	2017	2017	CD
hansen-can-2020	50	34	)	)	-RRB-
hansen-can-2020	50	35	,	,	,
hansen-can-2020	50	36	have	have	VBP
hansen-can-2020	50	37	portrayed	portray	VBN
hansen-can-2020	50	38	it	-PRON-	PRP
hansen-can-2020	50	39	.	.	.
hansen-can-2020	51	1	Let	let	VB
hansen-can-2020	51	2	us	-PRON-	PRP
hansen-can-2020	51	3	look	look	VB
hansen-can-2020	51	4	at	at	IN
hansen-can-2020	51	5	what	what	WP
hansen-can-2020	51	6	happens	happen	VBZ
hansen-can-2020	51	7	when	when	WRB
hansen-can-2020	51	8	smarter	smart	JJR
hansen-can-2020	51	9	and	and	CC
hansen-can-2020	51	10	more	more	RBR
hansen-can-2020	51	11	capable	capable	JJ
hansen-can-2020	51	12	minds	mind	NNS
hansen-can-2020	51	13	tackle	tackle	VBP
hansen-can-2020	51	14	the	the	DT
hansen-can-2020	51	15	problem	problem	NN
hansen-can-2020	51	16	of	of	IN
hansen-can-2020	51	17	classifying	classify	VBG
hansen-can-2020	51	18	mathematics	mathematic	NNS
hansen-can-2020	51	19	and	and	CC
hansen-can-2020	51	20	other	other	JJ
hansen-can-2020	51	21	highly	highly	RB
hansen-can-2020	51	22	technical	technical	JJ
hansen-can-2020	51	23	subjects	subject	NNS
hansen-can-2020	51	24	using	use	VBG
hansen-can-2020	51	25	advanced	advanced	JJ
hansen-can-2020	51	26	machine	machine	NN
hansen-can-2020	51	27	learning	learning	NN
hansen-can-2020	51	28	techniques	technique	NNS
hansen-can-2020	51	29	.	.	.
hansen-can-2020	52	1	Comment	comment	NN
hansen-can-2020	52	2	by	by	IN
hansen-can-2020	52	3	Mark	Mark	NNP
hansen-can-2020	52	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	52	5	:	:	:
hansen-can-2020	52	6	Perhaps	perhaps	RB
hansen-can-2020	52	7	a	a	DT
hansen-can-2020	52	8	different	different	JJ
hansen-can-2020	52	9	word	word	NN
hansen-can-2020	52	10	choice	choice	NN
hansen-can-2020	52	11	?	?	.
hansen-can-2020	53	1	Solving	solving	NN
hansen-can-2020	53	2	refers	refer	VBZ
hansen-can-2020	53	3	to	to	IN
hansen-can-2020	53	4	the	the	DT
hansen-can-2020	53	5	word	word	NN
hansen-can-2020	53	6	project	project	NN
hansen-can-2020	53	7	here	here	RB
hansen-can-2020	53	8	.	.	.
hansen-can-2020	54	1	Maybe	maybe	RB
hansen-can-2020	54	2	something	something	NN
hansen-can-2020	54	3	like	like	IN
hansen-can-2020	54	4	problem	problem	NN
hansen-can-2020	54	5	instead	instead	RB
hansen-can-2020	54	6	of	of	IN
hansen-can-2020	54	7	project	project	NN
hansen-can-2020	54	8	OR	or	CC
hansen-can-2020	54	9	supporting	support	VBG
hansen-can-2020	54	10	instead	instead	RB
hansen-can-2020	54	11	of	of	IN
hansen-can-2020	54	12	solving	solve	VBG
hansen-can-2020	54	13	?	?	.
hansen-can-2020	55	1	Comment	comment	NN
hansen-can-2020	55	2	by	by	IN
hansen-can-2020	55	3	Daniel	Daniel	NNP
hansen-can-2020	55	4	Johnson	Johnson	NNP
hansen-can-2020	55	5	:	:	:
hansen-can-2020	55	6	+1	+1	:
hansen-can-2020	55	7	Finding	find	VBG
hansen-can-2020	55	8	the	the	DT
hansen-can-2020	55	9	Right	Right	NNP
hansen-can-2020	55	10	Hammer	Hammer	NNP
hansen-can-2020	55	11	To	to	TO
hansen-can-2020	55	12	illustrate	illustrate	VB
hansen-can-2020	55	13	the	the	DT
hansen-can-2020	55	14	quest	quest	NN
hansen-can-2020	55	15	to	to	TO
hansen-can-2020	55	16	find	find	VB
hansen-can-2020	55	17	the	the	DT
hansen-can-2020	55	18	right	right	JJ
hansen-can-2020	55	19	hammer	hammer	NN
hansen-can-2020	55	20	I	-PRON-	PRP
hansen-can-2020	55	21	am	be	VBP
hansen-can-2020	55	22	going	go	VBG
hansen-can-2020	55	23	to	to	TO
hansen-can-2020	55	24	focus	focus	VB
hansen-can-2020	55	25	on	on	IN
hansen-can-2020	55	26	three	three	CD
hansen-can-2020	55	27	different	different	JJ
hansen-can-2020	55	28	projects	project	NNS
hansen-can-2020	55	29	that	that	WDT
hansen-can-2020	55	30	tackled	tackle	VBD
hansen-can-2020	55	31	the	the	DT
hansen-can-2020	55	32	automated	automate	VBN
hansen-can-2020	55	33	categorization	categorization	NN
hansen-can-2020	55	34	of	of	IN
hansen-can-2020	55	35	highly	highly	RB
hansen-can-2020	55	36	technical	technical	JJ
hansen-can-2020	55	37	content	content	NN
hansen-can-2020	55	38	,	,	,
hansen-can-2020	55	39	two	two	CD
hansen-can-2020	55	40	of	of	IN
hansen-can-2020	55	41	which	which	WDT
hansen-can-2020	55	42	also	also	RB
hansen-can-2020	55	43	attempted	attempt	VBD
hansen-can-2020	55	44	to	to	TO
hansen-can-2020	55	45	categorize	categorize	VB
hansen-can-2020	55	46	mathematical	mathematical	JJ
hansen-can-2020	55	47	content	content	NN
hansen-can-2020	55	48	and	and	CC
hansen-can-2020	55	49	one	one	CD
hansen-can-2020	55	50	that	that	WDT
hansen-can-2020	55	51	looked	look	VBD
hansen-can-2020	55	52	to	to	TO
hansen-can-2020	55	53	categorize	categorize	VB
hansen-can-2020	55	54	scholarly	scholarly	NNP
hansen-can-2020	55	55	works	work	NNS
hansen-can-2020	55	56	in	in	IN
hansen-can-2020	55	57	general	general	JJ
hansen-can-2020	55	58	.	.	.
hansen-can-2020	56	1	These	these	DT
hansen-can-2020	56	2	three	three	CD
hansen-can-2020	56	3	projects	project	NNS
hansen-can-2020	56	4	provide	provide	VBP
hansen-can-2020	56	5	examples	example	NNS
hansen-can-2020	56	6	of	of	IN
hansen-can-2020	56	7	many	many	JJ
hansen-can-2020	56	8	of	of	IN
hansen-can-2020	56	9	the	the	DT
hansen-can-2020	56	10	approaches	approach	NNS
hansen-can-2020	56	11	and	and	CC
hansen-can-2020	56	12	practices	practice	NNS
hansen-can-2020	56	13	employed	employ	VBN
hansen-can-2020	56	14	by	by	IN
hansen-can-2020	56	15	experts	expert	NNS
hansen-can-2020	56	16	in	in	IN
hansen-can-2020	56	17	automated	automated	JJ
hansen-can-2020	56	18	classification	classification	NN
hansen-can-2020	56	19	and	and	CC
hansen-can-2020	56	20	demonstrate	demonstrate	VB
hansen-can-2020	56	21	the	the	DT
hansen-can-2020	56	22	two	two	CD
hansen-can-2020	56	23	main	main	JJ
hansen-can-2020	56	24	paths	path	NNS
hansen-can-2020	56	25	that	that	WDT
hansen-can-2020	56	26	these	these	DT
hansen-can-2020	56	27	types	type	NNS
hansen-can-2020	56	28	of	of	IN
hansen-can-2020	56	29	projects	project	NNS
hansen-can-2020	56	30	follow	follow	VBP
hansen-can-2020	56	31	to	to	TO
hansen-can-2020	56	32	accomplish	accomplish	VB
hansen-can-2020	56	33	their	-PRON-	PRP$
hansen-can-2020	56	34	goals	goal	NNS
hansen-can-2020	56	35	.	.	.
hansen-can-2020	57	1	Since	since	IN
hansen-can-2020	57	2	we	-PRON-	PRP
hansen-can-2020	57	3	have	have	VBP
hansen-can-2020	57	4	been	be	VBN
hansen-can-2020	57	5	discussing	discuss	VBG
hansen-can-2020	57	6	mathematics	mathematic	NNS
hansen-can-2020	57	7	,	,	,
hansen-can-2020	57	8	let	let	VB
hansen-can-2020	57	9	us	-PRON-	PRP
hansen-can-2020	57	10	start	start	VB
hansen-can-2020	57	11	with	with	IN
hansen-can-2020	57	12	those	those	DT
hansen-can-2020	57	13	two	two	CD
hansen-can-2020	57	14	projects	project	NNS
hansen-can-2020	57	15	.	.	.
hansen-can-2020	58	1	Both	both	DT
hansen-can-2020	58	2	projects	project	NNS
hansen-can-2020	58	3	began	begin	VBD
hansen-can-2020	58	4	because	because	IN
hansen-can-2020	58	5	the	the	DT
hansen-can-2020	58	6	participants	participant	NNS
hansen-can-2020	58	7	were	be	VBD
hansen-can-2020	58	8	struggling	struggle	VBG
hansen-can-2020	58	9	to	to	TO
hansen-can-2020	58	10	categorize	categorize	VB
hansen-can-2020	58	11	mathematics	mathematic	NNS
hansen-can-2020	58	12	publications	publication	NNS
hansen-can-2020	58	13	so	so	IN
hansen-can-2020	58	14	they	-PRON-	PRP
hansen-can-2020	58	15	would	would	MD
hansen-can-2020	58	16	be	be	VB
hansen-can-2020	58	17	properly	properly	RB
hansen-can-2020	58	18	indexed	index	VBN
hansen-can-2020	58	19	and	and	CC
hansen-can-2020	58	20	searchable	searchable	JJ
hansen-can-2020	58	21	in	in	IN
hansen-can-2020	58	22	digital	digital	JJ
hansen-can-2020	58	23	mathematics	mathematic	NNS
hansen-can-2020	58	24	databases	database	NNS
hansen-can-2020	58	25	:	:	:
hansen-can-2020	58	26	the	the	DT
hansen-can-2020	58	27	Czech	Czech	NNP
hansen-can-2020	58	28	Digital	Digital	NNP
hansen-can-2020	58	29	Mathematics	Mathematics	NNP
hansen-can-2020	58	30	Library	Library	NNP
hansen-can-2020	58	31	(	(	-LRB-
hansen-can-2020	58	32	DML	DML	NNP
hansen-can-2020	58	33	-	-	HYPH
hansen-can-2020	58	34	CZ)[footnoteRef:5	CZ)[footnoteRef:5	NNP
hansen-can-2020	58	35	]	]	-RRB-
hansen-can-2020	58	36	and	and	CC
hansen-can-2020	58	37	NUMDAM[footnoteRef:6	numdam[footnoteref:6	RB
hansen-can-2020	58	38	]	]	-RRB-
hansen-can-2020	58	39	in	in	IN
hansen-can-2020	58	40	the	the	DT
hansen-can-2020	58	41	case	case	NN
hansen-can-2020	58	42	of	of	IN
hansen-can-2020	58	43	Radim	Radim	NNP
hansen-can-2020	58	44	Řehůřek	řehůřek	CD
hansen-can-2020	58	45	and	and	CC
hansen-can-2020	58	46	Petr	Petr	VBN
hansen-can-2020	58	47	Sojka	sojka	RB
hansen-can-2020	58	48	(	(	-LRB-
hansen-can-2020	58	49	Řehůřek	Řehůřek	NNP
hansen-can-2020	58	50	and	and	CC
hansen-can-2020	58	51	Sojka	sojka	RB
hansen-can-2020	58	52	2008	2008	CD
hansen-can-2020	58	53	)	)	-RRB-
hansen-can-2020	58	54	,	,	,
hansen-can-2020	58	55	and	and	CC
hansen-can-2020	58	56	Zentralblatt	Zentralblatt	NNP
hansen-can-2020	58	57	MATH	MATH	NNP
hansen-can-2020	58	58	(	(	-LRB-
hansen-can-2020	58	59	zbMath)[footnoteRef:7	zbmath)[footnoteref:7	FW
hansen-can-2020	58	60	]	]	-RRB-
hansen-can-2020	58	61	in	in	IN
hansen-can-2020	58	62	the	the	DT
hansen-can-2020	58	63	case	case	NN
hansen-can-2020	58	64	of	of	IN
hansen-can-2020	58	65	Simon	Simon	NNP
hansen-can-2020	58	66	Barthel	Barthel	NNP
hansen-can-2020	58	67	,	,	,
hansen-can-2020	58	68	Sascha	Sascha	NNP
hansen-can-2020	58	69	Tönnies	tönnie	NNS
hansen-can-2020	58	70	,	,	,
hansen-can-2020	58	71	and	and	CC
hansen-can-2020	58	72	Wolf	Wolf	NNP
hansen-can-2020	58	73	-	-	HYPH
hansen-can-2020	58	74	Tilo	Tilo	NNP
hansen-can-2020	58	75	Balke	Balke	NNP
hansen-can-2020	58	76	(	(	-LRB-
hansen-can-2020	58	77	Barthel	Barthel	NNP
hansen-can-2020	58	78	,	,	,
hansen-can-2020	58	79	Tönnies	Tönnies	NNP
hansen-can-2020	58	80	,	,	,
hansen-can-2020	58	81	and	and	CC
hansen-can-2020	58	82	Balke	Balke	NNP
hansen-can-2020	58	83	2013	2013	CD
hansen-can-2020	58	84	)	)	-RRB-
hansen-can-2020	58	85	.	.	.
hansen-can-2020	59	1	All	all	DT
hansen-can-2020	59	2	of	of	IN
hansen-can-2020	59	3	these	these	DT
hansen-can-2020	59	4	databases	database	NNS
hansen-can-2020	59	5	rely	rely	VBP
hansen-can-2020	59	6	on	on	IN
hansen-can-2020	59	7	the	the	DT
hansen-can-2020	59	8	aforementioned	aforementioned	JJ
hansen-can-2020	59	9	MSC[footnoteRef:8	MSC[footnoteRef:8	NNP
hansen-can-2020	59	10	]	]	-RRB-
hansen-can-2020	59	11	to	to	TO
hansen-can-2020	59	12	aid	aid	VB
hansen-can-2020	59	13	in	in	IN
hansen-can-2020	59	14	indexing	indexing	NN
hansen-can-2020	59	15	and	and	CC
hansen-can-2020	59	16	retrieval	retrieval	NN
hansen-can-2020	59	17	,	,	,
hansen-can-2020	59	18	and	and	CC
hansen-can-2020	59	19	so	so	RB
hansen-can-2020	59	20	their	-PRON-	PRP$
hansen-can-2020	59	21	goal	goal	NN
hansen-can-2020	59	22	was	be	VBD
hansen-can-2020	59	23	to	to	TO
hansen-can-2020	59	24	automate	automate	VB
hansen-can-2020	59	25	the	the	DT
hansen-can-2020	59	26	assignment	assignment	NN
hansen-can-2020	59	27	of	of	IN
hansen-can-2020	59	28	MSC	MSC	NNP
hansen-can-2020	59	29	values	value	NNS
hansen-can-2020	59	30	to	to	TO
hansen-can-2020	59	31	lower	lower	VB
hansen-can-2020	59	32	the	the	DT
hansen-can-2020	59	33	time	time	NN
hansen-can-2020	59	34	and	and	CC
hansen-can-2020	59	35	labor	labor	NN
hansen-can-2020	59	36	cost	cost	NN
hansen-can-2020	59	37	of	of	IN
hansen-can-2020	59	38	requiring	require	VBG
hansen-can-2020	59	39	humans	human	NNS
hansen-can-2020	59	40	to	to	TO
hansen-can-2020	59	41	do	do	VB
hansen-can-2020	59	42	this	this	DT
hansen-can-2020	59	43	task	task	NN
hansen-can-2020	59	44	.	.	.
hansen-can-2020	60	1	The	the	DT
hansen-can-2020	60	2	main	main	JJ
hansen-can-2020	60	3	differences	difference	NNS
hansen-can-2020	60	4	between	between	IN
hansen-can-2020	60	5	their	-PRON-	PRP$
hansen-can-2020	60	6	tasks	task	NNS
hansen-can-2020	60	7	related	relate	VBN
hansen-can-2020	60	8	to	to	IN
hansen-can-2020	60	9	the	the	DT
hansen-can-2020	60	10	number	number	NN
hansen-can-2020	60	11	of	of	IN
hansen-can-2020	60	12	documents	document	NNS
hansen-can-2020	60	13	they	-PRON-	PRP
hansen-can-2020	60	14	were	be	VBD
hansen-can-2020	60	15	working	work	VBG
hansen-can-2020	60	16	with	with	IN
hansen-can-2020	60	17	(	(	-LRB-
hansen-can-2020	60	18	thousands	thousand	NNS
hansen-can-2020	60	19	for	for	IN
hansen-can-2020	60	20	Řehůřek	Řehůřek	NNP
hansen-can-2020	60	21	and	and	CC
hansen-can-2020	60	22	Sojka	sojka	JJ
hansen-can-2020	60	23	and	and	CC
hansen-can-2020	60	24	millions	million	NNS
hansen-can-2020	60	25	for	for	IN
hansen-can-2020	60	26	Barthel	Barthel	NNP
hansen-can-2020	60	27	,	,	,
hansen-can-2020	60	28	Tönnies	Tönnies	NNP
hansen-can-2020	60	29	,	,	,
hansen-can-2020	60	30	and	and	CC
hansen-can-2020	60	31	Balke	Balke	NNP
hansen-can-2020	60	32	)	)	-RRB-
hansen-can-2020	60	33	,	,	,
hansen-can-2020	60	34	the	the	DT
hansen-can-2020	60	35	amount	amount	NN
hansen-can-2020	60	36	of	of	IN
hansen-can-2020	60	37	the	the	DT
hansen-can-2020	60	38	works	work	NNS
hansen-can-2020	60	39	available	available	JJ
hansen-can-2020	60	40	(	(	-LRB-
hansen-can-2020	60	41	full	full	JJ
hansen-can-2020	60	42	text	text	NN
hansen-can-2020	60	43	for	for	IN
hansen-can-2020	60	44	Řehůřek	Řehůřek	NNP
hansen-can-2020	60	45	and	and	CC
hansen-can-2020	60	46	Sojka	sojka	RB
hansen-can-2020	60	47	,	,	,
hansen-can-2020	60	48	and	and	CC
hansen-can-2020	60	49	titles	title	NNS
hansen-can-2020	60	50	,	,	,
hansen-can-2020	60	51	authors	author	NNS
hansen-can-2020	60	52	,	,	,
hansen-can-2020	60	53	and	and	CC
hansen-can-2020	60	54	abstracts	abstract	NNS
hansen-can-2020	60	55	for	for	IN
hansen-can-2020	60	56	Barthel	Barthel	NNP
hansen-can-2020	60	57	,	,	,
hansen-can-2020	60	58	Tönnies	Tönnies	NNP
hansen-can-2020	60	59	,	,	,
hansen-can-2020	60	60	and	and	CC
hansen-can-2020	60	61	Balke	Balke	NNP
hansen-can-2020	60	62	)	)	-RRB-
hansen-can-2020	60	63	,	,	,
hansen-can-2020	60	64	and	and	CC
hansen-can-2020	60	65	the	the	DT
hansen-can-2020	60	66	quality	quality	NN
hansen-can-2020	60	67	of	of	IN
hansen-can-2020	60	68	the	the	DT
hansen-can-2020	60	69	data	datum	NNS
hansen-can-2020	60	70	(	(	-LRB-
hansen-can-2020	60	71	mostly	mostly	RB
hansen-can-2020	60	72	OCR	ocr	NN
hansen-can-2020	60	73	scans	scan	NNS
hansen-can-2020	60	74	for	for	IN
hansen-can-2020	60	75	Řehůřek	Řehůřek	NNP
hansen-can-2020	60	76	and	and	CC
hansen-can-2020	60	77	Sojka	sojka	RB
hansen-can-2020	60	78	and	and	CC
hansen-can-2020	60	79	mostly	mostly	RB
hansen-can-2020	60	80	TeX	TeX	VBD
hansen-can-2020	60	81	for	for	IN
hansen-can-2020	60	82	Barthel	Barthel	NNP
hansen-can-2020	60	83	,	,	,
hansen-can-2020	60	84	Tönnies	Tönnies	NNP
hansen-can-2020	60	85	,	,	,
hansen-can-2020	60	86	and	and	CC
hansen-can-2020	60	87	Balke	Balke	NNP
hansen-can-2020	60	88	)	)	-RRB-
hansen-can-2020	60	89	.	.	.
hansen-can-2020	61	1	Even	even	RB
hansen-can-2020	61	2	with	with	IN
hansen-can-2020	61	3	these	these	DT
hansen-can-2020	61	4	differences	difference	NNS
hansen-can-2020	61	5	,	,	,
hansen-can-2020	61	6	both	both	DT
hansen-can-2020	61	7	projects	project	NNS
hansen-can-2020	61	8	took	take	VBD
hansen-can-2020	61	9	a	a	DT
hansen-can-2020	61	10	similar	similar	JJ
hansen-can-2020	61	11	approach	approach	NN
hansen-can-2020	61	12	,	,	,
hansen-can-2020	61	13	and	and	CC
hansen-can-2020	61	14	it	-PRON-	PRP
hansen-can-2020	61	15	is	be	VBZ
hansen-can-2020	61	16	the	the	DT
hansen-can-2020	61	17	first	first	JJ
hansen-can-2020	61	18	of	of	IN
hansen-can-2020	61	19	the	the	DT
hansen-can-2020	61	20	two	two	CD
hansen-can-2020	61	21	main	main	JJ
hansen-can-2020	61	22	pathways	pathway	NNS
hansen-can-2020	61	23	toward	toward	IN
hansen-can-2020	61	24	classification	classification	NN
hansen-can-2020	61	25	I	-PRON-	PRP
hansen-can-2020	61	26	spoke	speak	VBD
hansen-can-2020	61	27	of	of	IN
hansen-can-2020	61	28	earlier	early	RBR
hansen-can-2020	61	29	:	:	:
hansen-can-2020	61	30	using	use	VBG
hansen-can-2020	61	31	a	a	DT
hansen-can-2020	61	32	predetermined	predetermined	JJ
hansen-can-2020	61	33	taxonomy	taxonomy	NN
hansen-can-2020	61	34	and	and	CC
hansen-can-2020	61	35	a	a	DT
hansen-can-2020	61	36	set	set	NN
hansen-can-2020	61	37	of	of	IN
hansen-can-2020	61	38	pre	pre	JJ
hansen-can-2020	61	39	-	-	JJ
hansen-can-2020	61	40	categorized	categorized	JJ
hansen-can-2020	61	41	data	datum	NNS
hansen-can-2020	61	42	to	to	TO
hansen-can-2020	61	43	build	build	VB
hansen-can-2020	61	44	a	a	DT
hansen-can-2020	61	45	machine	machine	NN
hansen-can-2020	61	46	learning	learn	VBG
hansen-can-2020	61	47	categorizer	categorizer	NN
hansen-can-2020	61	48	.	.	.
hansen-can-2020	62	1	[	[	-LRB-
hansen-can-2020	62	2	5	5	CD
hansen-can-2020	62	3	:	:	:
hansen-can-2020	62	4	https://dml.cz/	https://dml.cz/	NNP
hansen-can-2020	62	5	]	]	-RRB-
hansen-can-2020	62	6	[	[	-LRB-
hansen-can-2020	62	7	6	6	CD
hansen-can-2020	62	8	:	:	:
hansen-can-2020	62	9	http://www.numdam.org/	http://www.numdam.org/	NN
hansen-can-2020	62	10	]	]	-RRB-
hansen-can-2020	62	11	[	[	-LRB-
hansen-can-2020	62	12	7	7	CD
hansen-can-2020	62	13	:	:	:
hansen-can-2020	62	14	https://zbmath.org/	https://zbmath.org/	LS
hansen-can-2020	62	15	]	]	-RRB-
hansen-can-2020	62	16	[	[	-LRB-
hansen-can-2020	62	17	8	8	CD
hansen-can-2020	62	18	:	:	:
hansen-can-2020	62	19	Mathematical	Mathematical	NNP
hansen-can-2020	62	20	Subject	Subject	NNP
hansen-can-2020	62	21	Classification	Classification	NNP
hansen-can-2020	62	22	(	(	-LRB-
hansen-can-2020	62	23	MSC	MSC	NNP
hansen-can-2020	62	24	)	)	-RRB-
hansen-can-2020	62	25	values	value	NNS
hansen-can-2020	62	26	in	in	IN
hansen-can-2020	62	27	MathSciNet	MathSciNet	NNP
hansen-can-2020	62	28	and	and	CC
hansen-can-2020	62	29	zbMath	zbMath	NNP
hansen-can-2020	62	30	are	be	VBP
hansen-can-2020	62	31	a	a	DT
hansen-can-2020	62	32	particularly	particularly	RB
hansen-can-2020	62	33	interesting	interesting	JJ
hansen-can-2020	62	34	categorization	categorization	NN
hansen-can-2020	62	35	set	set	VBN
hansen-can-2020	62	36	to	to	TO
hansen-can-2020	62	37	work	work	VB
hansen-can-2020	62	38	with	with	IN
hansen-can-2020	62	39	as	as	IN
hansen-can-2020	62	40	they	-PRON-	PRP
hansen-can-2020	62	41	are	be	VBP
hansen-can-2020	62	42	assigned	assign	VBN
hansen-can-2020	62	43	and	and	CC
hansen-can-2020	62	44	reviewed	review	VBN
hansen-can-2020	62	45	by	by	IN
hansen-can-2020	62	46	a	a	DT
hansen-can-2020	62	47	subject	subject	JJ
hansen-can-2020	62	48	area	area	NN
hansen-can-2020	62	49	expert	expert	NN
hansen-can-2020	62	50	editor	editor	NN
hansen-can-2020	62	51	and	and	CC
hansen-can-2020	62	52	an	an	DT
hansen-can-2020	62	53	active	active	JJ
hansen-can-2020	62	54	researcher	researcher	NN
hansen-can-2020	62	55	in	in	IN
hansen-can-2020	62	56	the	the	DT
hansen-can-2020	62	57	same	same	JJ
hansen-can-2020	62	58	,	,	,
hansen-can-2020	62	59	or	or	CC
hansen-can-2020	62	60	closely	closely	RB
hansen-can-2020	62	61	related	relate	VBN
hansen-can-2020	62	62	,	,	,
hansen-can-2020	62	63	subfield	subfield	JJ
hansen-can-2020	62	64	as	as	IN
hansen-can-2020	62	65	the	the	DT
hansen-can-2020	62	66	article	article	NN
hansen-can-2020	62	67	’s	’s	POS
hansen-can-2020	62	68	content	content	NN
hansen-can-2020	62	69	before	before	IN
hansen-can-2020	62	70	they	-PRON-	PRP
hansen-can-2020	62	71	are	be	VBP
hansen-can-2020	62	72	published	publish	VBN
hansen-can-2020	62	73	.	.	.
hansen-can-2020	63	1	This	this	DT
hansen-can-2020	63	2	multi	multi	JJ
hansen-can-2020	63	3	-	-	JJ
hansen-can-2020	63	4	step	step	JJ
hansen-can-2020	63	5	process	process	NN
hansen-can-2020	63	6	of	of	IN
hansen-can-2020	63	7	review	review	NN
hansen-can-2020	63	8	yields	yield	NNS
hansen-can-2020	63	9	a	a	DT
hansen-can-2020	63	10	built	build	VBN
hansen-can-2020	63	11	-	-	HYPH
hansen-can-2020	63	12	in	in	RP
hansen-can-2020	63	13	accuracy	accuracy	NN
hansen-can-2020	63	14	check	check	NN
hansen-can-2020	63	15	for	for	IN
hansen-can-2020	63	16	the	the	DT
hansen-can-2020	63	17	categorization	categorization	NN
hansen-can-2020	63	18	.	.	.
hansen-can-2020	63	19	]	]	-RRB-
hansen-can-2020	64	1	In	in	IN
hansen-can-2020	64	2	the	the	DT
hansen-can-2020	64	3	end	end	NN
hansen-can-2020	64	4	,	,	,
hansen-can-2020	64	5	while	while	IN
hansen-can-2020	64	6	both	both	DT
hansen-can-2020	64	7	projects	project	NNS
hansen-can-2020	64	8	determined	determine	VBD
hansen-can-2020	64	9	that	that	IN
hansen-can-2020	64	10	the	the	DT
hansen-can-2020	64	11	use	use	NN
hansen-can-2020	64	12	of	of	IN
hansen-can-2020	64	13	Support	Support	NNP
hansen-can-2020	64	14	Vector	Vector	NNP
hansen-can-2020	64	15	Machines	Machines	NNPS
hansen-can-2020	64	16	(	(	-LRB-
hansen-can-2020	64	17	Gandhi	Gandhi	NNP
hansen-can-2020	64	18	2018)[footnoteRef:9	2018)[footnoteRef:9	NNP
hansen-can-2020	64	19	]	]	-RRB-
hansen-can-2020	64	20	provided	provide	VBD
hansen-can-2020	64	21	the	the	DT
hansen-can-2020	64	22	best	good	JJS
hansen-can-2020	64	23	categorization	categorization	NN
hansen-can-2020	64	24	results	result	NNS
hansen-can-2020	64	25	,	,	,
hansen-can-2020	64	26	their	-PRON-	PRP$
hansen-can-2020	64	27	implementations	implementation	NNS
hansen-can-2020	64	28	were	be	VBD
hansen-can-2020	64	29	different	different	JJ
hansen-can-2020	64	30	.	.	.
hansen-can-2020	65	1	The	the	DT
hansen-can-2020	65	2	Řehůřek	Řehůřek	NNP
hansen-can-2020	65	3	and	and	CC
hansen-can-2020	65	4	Sojka	sojka	RB
hansen-can-2020	65	5	SVMs	svm	NNS
hansen-can-2020	65	6	were	be	VBD
hansen-can-2020	65	7	trained	train	VBN
hansen-can-2020	65	8	with	with	IN
hansen-can-2020	65	9	terms	term	NNS
hansen-can-2020	65	10	weighted	weight	VBN
hansen-can-2020	65	11	using	use	VBG
hansen-can-2020	65	12	augmented	augment	VBN
hansen-can-2020	65	13	term	term	NN
hansen-can-2020	65	14	frequency[footnoteRef:10	frequency[footnoteref:10	NN
hansen-can-2020	65	15	]	]	-RRB-
hansen-can-2020	65	16	and	and	CC
hansen-can-2020	65	17	dynamic	dynamic	JJ
hansen-can-2020	65	18	decision	decision	NN
hansen-can-2020	65	19	threshold[footnoteRef:11	threshold[footnoteref:11	NN
hansen-can-2020	65	20	]	]	-RRB-
hansen-can-2020	65	21	selection	selection	NN
hansen-can-2020	65	22	using	use	VBG
hansen-can-2020	65	23	s	s	NNPS
hansen-can-2020	65	24	-	-	HYPH
hansen-can-2020	65	25	cut[footnoteRef:12	cut[footnoteRef:12	NNP
hansen-can-2020	65	26	]	]	-RRB-
hansen-can-2020	65	27	(	(	-LRB-
hansen-can-2020	65	28	Řehůřek	Řehůřek	NNP
hansen-can-2020	65	29	and	and	CC
hansen-can-2020	65	30	Sojka	sojka	RB
hansen-can-2020	65	31	2008	2008	CD
hansen-can-2020	65	32	,	,	,
hansen-can-2020	65	33	549	549	CD
hansen-can-2020	65	34	)	)	-RRB-
hansen-can-2020	65	35	and	and	CC
hansen-can-2020	65	36	Barthel	Barthel	NNP
hansen-can-2020	65	37	,	,	,
hansen-can-2020	65	38	Tönnies	Tönnies	NNP
hansen-can-2020	65	39	,	,	,
hansen-can-2020	65	40	and	and	CC
hansen-can-2020	65	41	Balke	Balke	NNP
hansen-can-2020	65	42	's	's	POS
hansen-can-2020	65	43	with	with	IN
hansen-can-2020	65	44	term	term	NN
hansen-can-2020	65	45	weighting	weight	VBG
hansen-can-2020	65	46	using	use	VBG
hansen-can-2020	65	47	term	term	NN
hansen-can-2020	65	48	frequency	frequency	NN
hansen-can-2020	65	49	–	–	:
hansen-can-2020	65	50	inverse	inverse	NN
hansen-can-2020	65	51	document	document	NN
hansen-can-2020	65	52	frequency[footnoteRef:13	frequency[footnoteRef:13	NNP
hansen-can-2020	65	53	]	]	-RRB-
hansen-can-2020	65	54	and	and	CC
hansen-can-2020	65	55	Euclidean	Euclidean	NNP
hansen-can-2020	65	56	normalization[footnoteRef:14	normalization[footnoteRef:14	''
hansen-can-2020	65	57	]	]	-RRB-
hansen-can-2020	65	58	(	(	-LRB-
hansen-can-2020	65	59	Barthel	Barthel	NNP
hansen-can-2020	65	60	,	,	,
hansen-can-2020	65	61	Tönnies	Tönnies	NNP
hansen-can-2020	65	62	,	,	,
hansen-can-2020	65	63	and	and	CC
hansen-can-2020	65	64	Balke	Balke	NNP
hansen-can-2020	65	65	2013	2013	CD
hansen-can-2020	65	66	,	,	,
hansen-can-2020	65	67	88	88	CD
hansen-can-2020	65	68	)	)	-RRB-
hansen-can-2020	65	69	,	,	,
hansen-can-2020	65	70	but	but	CC
hansen-can-2020	65	71	the	the	DT
hansen-can-2020	65	72	main	main	JJ
hansen-can-2020	65	73	difference	difference	NN
hansen-can-2020	65	74	was	be	VBD
hansen-can-2020	65	75	how	how	WRB
hansen-can-2020	65	76	they	-PRON-	PRP
hansen-can-2020	65	77	handled	handle	VBD
hansen-can-2020	65	78	formulae	formulae	NNP
hansen-can-2020	65	79	.	.	.
hansen-can-2020	66	1	In	in	IN
hansen-can-2020	66	2	particular	particular	JJ
hansen-can-2020	66	3	the	the	DT
hansen-can-2020	66	4	Barthel	Barthel	NNP
hansen-can-2020	66	5	,	,	,
hansen-can-2020	66	6	Tönnies	Tönnies	NNP
hansen-can-2020	66	7	,	,	,
hansen-can-2020	66	8	and	and	CC
hansen-can-2020	66	9	Balke	Balke	NNP
hansen-can-2020	66	10	group	group	NN
hansen-can-2020	66	11	split	split	VBD
hansen-can-2020	66	12	their	-PRON-	PRP$
hansen-can-2020	66	13	corpus	corpus	NN
hansen-can-2020	66	14	into	into	IN
hansen-can-2020	66	15	words	word	NNS
hansen-can-2020	66	16	and	and	CC
hansen-can-2020	66	17	formulae	formula	NNS
hansen-can-2020	66	18	and	and	CC
hansen-can-2020	66	19	mapped	map	VBD
hansen-can-2020	66	20	them	-PRON-	PRP
hansen-can-2020	66	21	to	to	TO
hansen-can-2020	66	22	separate	separate	JJ
hansen-can-2020	66	23	vectors	vector	NNS
hansen-can-2020	66	24	which	which	WDT
hansen-can-2020	66	25	were	be	VBD
hansen-can-2020	66	26	then	then	RB
hansen-can-2020	66	27	merged	merge	VBN
hansen-can-2020	66	28	together	together	RB
hansen-can-2020	66	29	for	for	IN
hansen-can-2020	66	30	a	a	DT
hansen-can-2020	66	31	combined	combine	VBN
hansen-can-2020	66	32	vector	vector	NN
hansen-can-2020	66	33	used	use	VBN
hansen-can-2020	66	34	for	for	IN
hansen-can-2020	66	35	categorization	categorization	NN
hansen-can-2020	66	36	.	.	.
hansen-can-2020	67	1	Řehůřek	Řehůřek	NNP
hansen-can-2020	67	2	and	and	CC
hansen-can-2020	67	3	Sojka	Sojka	NNP
hansen-can-2020	67	4	did	do	VBD
hansen-can-2020	67	5	not	not	RB
hansen-can-2020	67	6	differentiate	differentiate	VB
hansen-can-2020	67	7	between	between	IN
hansen-can-2020	67	8	words	word	NNS
hansen-can-2020	67	9	and	and	CC
hansen-can-2020	67	10	formulae	formula	NNS
hansen-can-2020	67	11	in	in	IN
hansen-can-2020	67	12	their	-PRON-	PRP$
hansen-can-2020	67	13	corpus	corpus	NN
hansen-can-2020	67	14	,	,	,
hansen-can-2020	67	15	and	and	CC
hansen-can-2020	67	16	they	-PRON-	PRP
hansen-can-2020	67	17	did	do	VBD
hansen-can-2020	67	18	note	note	VB
hansen-can-2020	67	19	that	that	IN
hansen-can-2020	67	20	their	-PRON-	PRP$
hansen-can-2020	67	21	OCR	ocr	NN
hansen-can-2020	67	22	scans	scan	NNS
hansen-can-2020	67	23	’	’	POS
hansen-can-2020	67	24	poor	poor	JJ
hansen-can-2020	67	25	handling	handling	NN
hansen-can-2020	67	26	of	of	IN
hansen-can-2020	67	27	formulae	formula	NNS
hansen-can-2020	67	28	could	could	MD
hansen-can-2020	67	29	have	have	VB
hansen-can-2020	67	30	hindered	hinder	VBN
hansen-can-2020	67	31	their	-PRON-	PRP$
hansen-can-2020	67	32	results	result	NNS
hansen-can-2020	67	33	(	(	-LRB-
hansen-can-2020	67	34	Řehůřek	Řehůřek	NNP
hansen-can-2020	67	35	and	and	CC
hansen-can-2020	67	36	Sojka	sojka	RB
hansen-can-2020	67	37	2008	2008	CD
hansen-can-2020	67	38	,	,	,
hansen-can-2020	67	39	555	555	CD
hansen-can-2020	67	40	)	)	-RRB-
hansen-can-2020	67	41	.	.	.
hansen-can-2020	68	1	In	in	IN
hansen-can-2020	68	2	the	the	DT
hansen-can-2020	68	3	end	end	NN
hansen-can-2020	68	4	,	,	,
hansen-can-2020	68	5	not	not	RB
hansen-can-2020	68	6	having	have	VBG
hansen-can-2020	68	7	the	the	DT
hansen-can-2020	68	8	ability	ability	NN
hansen-can-2020	68	9	to	to	TO
hansen-can-2020	68	10	handle	handle	VB
hansen-can-2020	68	11	formulae	formulae	NNP
hansen-can-2020	68	12	separately	separately	RB
hansen-can-2020	68	13	did	do	VBD
hansen-can-2020	68	14	not	not	RB
hansen-can-2020	68	15	seem	seem	VB
hansen-can-2020	68	16	to	to	TO
hansen-can-2020	68	17	matter	matter	VB
hansen-can-2020	68	18	as	as	IN
hansen-can-2020	68	19	Řehůřek	Řehůřek	NNP
hansen-can-2020	68	20	and	and	CC
hansen-can-2020	68	21	Sojka	Sojka	NNP
hansen-can-2020	68	22	claimed	claim	VBD
hansen-can-2020	68	23	microaveraged	microaverage	VBD
hansen-can-2020	68	24	F₁	F₁	NNP
hansen-can-2020	68	25	scores	score	NNS
hansen-can-2020	68	26	of	of	IN
hansen-can-2020	68	27	89.03	89.03	CD
hansen-can-2020	68	28	%	%	NN
hansen-can-2020	68	29	(	(	-LRB-
hansen-can-2020	68	30	Řehůřek	Řehůřek	NNP
hansen-can-2020	68	31	and	and	CC
hansen-can-2020	68	32	Sojka	sojka	RB
hansen-can-2020	68	33	2008	2008	CD
hansen-can-2020	68	34	,	,	,
hansen-can-2020	68	35	549	549	CD
hansen-can-2020	68	36	)	)	-RRB-
hansen-can-2020	68	37	when	when	WRB
hansen-can-2020	68	38	classifying	classify	VBG
hansen-can-2020	68	39	the	the	DT
hansen-can-2020	68	40	top	top	JJ
hansen-can-2020	68	41	level	level	NN
hansen-can-2020	68	42	MSC	MSC	NNP
hansen-can-2020	68	43	category	category	NN
hansen-can-2020	68	44	with	with	IN
hansen-can-2020	68	45	their	-PRON-	PRP$
hansen-can-2020	68	46	best	good	JJS
hansen-can-2020	68	47	performing	perform	VBG
hansen-can-2020	68	48	SVM	SVM	NNP
hansen-can-2020	68	49	.	.	.
hansen-can-2020	69	1	When	when	WRB
hansen-can-2020	69	2	this	this	DT
hansen-can-2020	69	3	is	be	VBZ
hansen-can-2020	69	4	compared	compare	VBN
hansen-can-2020	69	5	to	to	IN
hansen-can-2020	69	6	the	the	DT
hansen-can-2020	69	7	microaveraged	microaverage	VBN
hansen-can-2020	69	8	F₁	F₁	NNP
hansen-can-2020	69	9	of	of	IN
hansen-can-2020	69	10	67.3	67.3	CD
hansen-can-2020	69	11	%	%	NN
hansen-can-2020	69	12	obtained	obtain	VBN
hansen-can-2020	69	13	by	by	IN
hansen-can-2020	69	14	Barthel	Barthel	NNP
hansen-can-2020	69	15	,	,	,
hansen-can-2020	69	16	Tönnies	Tönnies	NNP
hansen-can-2020	69	17	,	,	,
hansen-can-2020	69	18	and	and	CC
hansen-can-2020	69	19	Balke	Balke	NNP
hansen-can-2020	69	20	(	(	-LRB-
hansen-can-2020	69	21	Barthel	Barthel	NNP
hansen-can-2020	69	22	,	,	,
hansen-can-2020	69	23	Tönnies	Tönnies	NNP
hansen-can-2020	69	24	,	,	,
hansen-can-2020	69	25	and	and	CC
hansen-can-2020	69	26	Balke	Balke	NNP
hansen-can-2020	69	27	2013	2013	CD
hansen-can-2020	69	28	,	,	,
hansen-can-2020	69	29	88	88	CD
hansen-can-2020	69	30	)	)	-RRB-
hansen-can-2020	69	31	,	,	,
hansen-can-2020	69	32	it	-PRON-	PRP
hansen-can-2020	69	33	would	would	MD
hansen-can-2020	69	34	seem	seem	VB
hansen-can-2020	69	35	that	that	IN
hansen-can-2020	69	36	either	either	CC
hansen-can-2020	69	37	Řehůřek	Řehůřek	NNP
hansen-can-2020	69	38	’s	’s	,
hansen-can-2020	69	39	and	and	CC
hansen-can-2020	69	40	Sojka	sojka	RB
hansen-can-2020	69	41	’s	’	VBZ
hansen-can-2020	69	42	implementation	implementation	NN
hansen-can-2020	69	43	of	of	IN
hansen-can-2020	69	44	SVMs	svm	NNS
hansen-can-2020	69	45	or	or	CC
hansen-can-2020	69	46	their	-PRON-	PRP$
hansen-can-2020	69	47	access	access	NN
hansen-can-2020	69	48	to	to	IN
hansen-can-2020	69	49	full	full	JJ
hansen-can-2020	69	50	-	-	HYPH
hansen-can-2020	69	51	text	text	NN
hansen-can-2020	69	52	led	lead	VBD
hansen-can-2020	69	53	to	to	IN
hansen-can-2020	69	54	a	a	DT
hansen-can-2020	69	55	clear	clear	JJ
hansen-can-2020	69	56	advantage	advantage	NN
hansen-can-2020	69	57	.	.	.
hansen-can-2020	70	1	This	this	DT
hansen-can-2020	70	2	advantage	advantage	NN
hansen-can-2020	70	3	becomes	become	VBZ
hansen-can-2020	70	4	less	less	RBR
hansen-can-2020	70	5	clear	clear	JJ
hansen-can-2020	70	6	when	when	WRB
hansen-can-2020	70	7	one	one	PRP
hansen-can-2020	70	8	takes	take	VBZ
hansen-can-2020	70	9	into	into	IN
hansen-can-2020	70	10	account	account	NN
hansen-can-2020	70	11	that	that	IN
hansen-can-2020	70	12	Řehůřek	Řehůřek	NNP
hansen-can-2020	70	13	and	and	CC
hansen-can-2020	70	14	Sojka	Sojka	NNP
hansen-can-2020	70	15	were	be	VBD
hansen-can-2020	70	16	only	only	RB
hansen-can-2020	70	17	working	work	VBG
hansen-can-2020	70	18	with	with	IN
hansen-can-2020	70	19	top	top	JJ
hansen-can-2020	70	20	level	level	NN
hansen-can-2020	70	21	MSCs	msc	NNS
hansen-can-2020	70	22	where	where	WRB
hansen-can-2020	70	23	they	-PRON-	PRP
hansen-can-2020	70	24	had	have	VBD
hansen-can-2020	70	25	at	at	RB
hansen-can-2020	70	26	least	least	RBS
hansen-can-2020	70	27	30	30	CD
hansen-can-2020	70	28	(	(	-LRB-
hansen-can-2020	70	29	60	60	CD
hansen-can-2020	70	30	in	in	IN
hansen-can-2020	70	31	the	the	DT
hansen-can-2020	70	32	case	case	NN
hansen-can-2020	70	33	of	of	IN
hansen-can-2020	70	34	their	-PRON-	PRP$
hansen-can-2020	70	35	best	good	JJS
hansen-can-2020	70	36	result	result	NN
hansen-can-2020	70	37	)	)	-RRB-
hansen-can-2020	70	38	articles	article	NNS
hansen-can-2020	70	39	,	,	,
hansen-can-2020	70	40	and	and	CC
hansen-can-2020	70	41	their	-PRON-	PRP$
hansen-can-2020	70	42	limited	limited	JJ
hansen-can-2020	70	43	corpus	corpus	NN
hansen-can-2020	70	44	meant	mean	VBD
hansen-can-2020	70	45	that	that	IN
hansen-can-2020	70	46	many	many	JJ
hansen-can-2020	70	47	top	top	JJ
hansen-can-2020	70	48	-	-	HYPH
hansen-can-2020	70	49	level	level	NN
hansen-can-2020	70	50	MSC	MSC	NNP
hansen-can-2020	70	51	categories	category	NNS
hansen-can-2020	70	52	would	would	MD
hansen-can-2020	70	53	not	not	RB
hansen-can-2020	70	54	have	have	VB
hansen-can-2020	70	55	been	be	VBN
hansen-can-2020	70	56	included	include	VBN
hansen-can-2020	70	57	.	.	.
hansen-can-2020	71	1	Looking	look	VBG
hansen-can-2020	71	2	at	at	IN
hansen-can-2020	71	3	the	the	DT
hansen-can-2020	71	4	work	work	NN
hansen-can-2020	71	5	done	do	VBN
hansen-can-2020	71	6	by	by	IN
hansen-can-2020	71	7	Barthel	Barthel	NNP
hansen-can-2020	71	8	,	,	,
hansen-can-2020	71	9	Tönnies	Tönnies	NNP
hansen-can-2020	71	10	,	,	,
hansen-can-2020	71	11	and	and	CC
hansen-can-2020	71	12	Balke	Balke	NNP
hansen-can-2020	71	13	makes	make	VBZ
hansen-can-2020	71	14	it	-PRON-	PRP
hansen-can-2020	71	15	clear	clear	JJ
hansen-can-2020	71	16	that	that	IN
hansen-can-2020	71	17	these	these	DT
hansen-can-2020	71	18	less	less	RBR
hansen-can-2020	71	19	common	common	JJ
hansen-can-2020	71	20	MSC	MSC	NNP
hansen-can-2020	71	21	categories	category	NNS
hansen-can-2020	71	22	such	such	JJ
hansen-can-2020	71	23	as	as	IN
hansen-can-2020	71	24	K	K	NNP
hansen-can-2020	71	25	-	-	HYPH
hansen-can-2020	71	26	Theory	Theory	NNP
hansen-can-2020	71	27	or	or	CC
hansen-can-2020	71	28	Potential	Potential	NNP
hansen-can-2020	71	29	Theory	Theory	NNP
hansen-can-2020	71	30	,	,	,
hansen-can-2020	71	31	for	for	IN
hansen-can-2020	71	32	which	which	WDT
hansen-can-2020	71	33	Barthel	Barthel	NNP
hansen-can-2020	71	34	,	,	,
hansen-can-2020	71	35	Tönnies	Tönnies	NNP
hansen-can-2020	71	36	,	,	,
hansen-can-2020	71	37	and	and	CC
hansen-can-2020	71	38	Balke	Balke	NNP
hansen-can-2020	71	39	achieved	achieve	VBD
hansen-can-2020	71	40	microaveraged	microaverage	VBD
hansen-can-2020	71	41	F₁	F₁	NNP
hansen-can-2020	71	42	measures	measure	NNS
hansen-can-2020	71	43	of	of	IN
hansen-can-2020	71	44	18.2	18.2	CD
hansen-can-2020	71	45	%	%	NN
hansen-can-2020	71	46	and	and	CC
hansen-can-2020	71	47	24	24	CD
hansen-can-2020	71	48	%	%	NN
hansen-can-2020	71	49	respectively	respectively	RB
hansen-can-2020	71	50	,	,	,
hansen-can-2020	71	51	have	have	VBP
hansen-can-2020	71	52	a	a	DT
hansen-can-2020	71	53	large	large	JJ
hansen-can-2020	71	54	impact	impact	NN
hansen-can-2020	71	55	on	on	IN
hansen-can-2020	71	56	the	the	DT
hansen-can-2020	71	57	overall	overall	JJ
hansen-can-2020	71	58	effectiveness	effectiveness	NN
hansen-can-2020	71	59	of	of	IN
hansen-can-2020	71	60	the	the	DT
hansen-can-2020	71	61	automated	automated	JJ
hansen-can-2020	71	62	categorization	categorization	NN
hansen-can-2020	71	63	.	.	.
hansen-can-2020	72	1	Remember	remember	VB
hansen-can-2020	72	2	,	,	,
hansen-can-2020	72	3	this	this	DT
hansen-can-2020	72	4	is	be	VBZ
hansen-can-2020	72	5	only	only	RB
hansen-can-2020	72	6	for	for	IN
hansen-can-2020	72	7	the	the	DT
hansen-can-2020	72	8	top	top	JJ
hansen-can-2020	72	9	level	level	NN
hansen-can-2020	72	10	of	of	IN
hansen-can-2020	72	11	MSC	MSC	NNP
hansen-can-2020	72	12	codes	code	NNS
hansen-can-2020	72	13	,	,	,
hansen-can-2020	72	14	and	and	CC
hansen-can-2020	72	15	the	the	DT
hansen-can-2020	72	16	work	work	NN
hansen-can-2020	72	17	of	of	IN
hansen-can-2020	72	18	Barthel	Barthel	NNP
hansen-can-2020	72	19	,	,	,
hansen-can-2020	72	20	Tönnies	Tönnies	NNP
hansen-can-2020	72	21	,	,	,
hansen-can-2020	72	22	and	and	CC
hansen-can-2020	72	23	Balke	Balke	NNP
hansen-can-2020	72	24	suggests	suggest	VBZ
hansen-can-2020	72	25	it	-PRON-	PRP
hansen-can-2020	72	26	would	would	MD
hansen-can-2020	72	27	get	get	VB
hansen-can-2020	72	28	worse	bad	JJR
hansen-can-2020	72	29	when	when	WRB
hansen-can-2020	72	30	trying	try	VBG
hansen-can-2020	72	31	to	to	TO
hansen-can-2020	72	32	apply	apply	VB
hansen-can-2020	72	33	the	the	DT
hansen-can-2020	72	34	second	second	JJ
hansen-can-2020	72	35	and	and	CC
hansen-can-2020	72	36	third	third	JJ
hansen-can-2020	72	37	level	level	NN
hansen-can-2020	72	38	for	for	IN
hansen-can-2020	72	39	full	full	JJ
hansen-can-2020	72	40	MSC	MSC	NNP
hansen-can-2020	72	41	categorization	categorization	NN
hansen-can-2020	72	42	to	to	IN
hansen-can-2020	72	43	these	these	DT
hansen-can-2020	72	44	less	less	RBR
hansen-can-2020	72	45	-	-	HYPH
hansen-can-2020	72	46	common	common	JJ
hansen-can-2020	72	47	categories	category	NNS
hansen-can-2020	72	48	.	.	.
hansen-can-2020	73	1	This	this	DT
hansen-can-2020	73	2	leads	lead	VBZ
hansen-can-2020	73	3	me	-PRON-	PRP
hansen-can-2020	73	4	to	to	TO
hansen-can-2020	73	5	believe	believe	VB
hansen-can-2020	73	6	that	that	IN
hansen-can-2020	73	7	in	in	IN
hansen-can-2020	73	8	the	the	DT
hansen-can-2020	73	9	case	case	NN
hansen-can-2020	73	10	of	of	IN
hansen-can-2020	73	11	categorizing	categorize	VBG
hansen-can-2020	73	12	highly	highly	RB
hansen-can-2020	73	13	technical	technical	JJ
hansen-can-2020	73	14	mathematical	mathematical	JJ
hansen-can-2020	73	15	works	work	NNS
hansen-can-2020	73	16	to	to	IN
hansen-can-2020	73	17	an	an	DT
hansen-can-2020	73	18	existing	exist	VBG
hansen-can-2020	73	19	taxonomy	taxonomy	NN
hansen-can-2020	73	20	,	,	,
hansen-can-2020	73	21	people	people	NNS
hansen-can-2020	73	22	have	have	VBP
hansen-can-2020	73	23	come	come	VBN
hansen-can-2020	73	24	close	close	RB
hansen-can-2020	73	25	to	to	IN
hansen-can-2020	73	26	identifying	identify	VBG
hansen-can-2020	73	27	the	the	DT
hansen-can-2020	73	28	overall	overall	JJ
hansen-can-2020	73	29	size	size	NN
hansen-can-2020	73	30	of	of	IN
hansen-can-2020	73	31	the	the	DT
hansen-can-2020	73	32	machine	machine	NN
hansen-can-2020	73	33	learning	learning	NN
hansen-can-2020	73	34	hammer	hammer	NNP
hansen-can-2020	73	35	,	,	,
hansen-can-2020	73	36	but	but	CC
hansen-can-2020	73	37	are	be	VBP
hansen-can-2020	73	38	still	still	RB
hansen-can-2020	73	39	a	a	DT
hansen-can-2020	73	40	long	long	JJ
hansen-can-2020	73	41	way	way	NN
hansen-can-2020	73	42	away	away	RB
hansen-can-2020	73	43	from	from	IN
hansen-can-2020	73	44	finding	find	VBG
hansen-can-2020	73	45	the	the	DT
hansen-can-2020	73	46	right	right	JJ
hansen-can-2020	73	47	match	match	NN
hansen-can-2020	73	48	for	for	IN
hansen-can-2020	73	49	the	the	DT
hansen-can-2020	73	50	categorization	categorization	NN
hansen-can-2020	73	51	nail	nail	NN
hansen-can-2020	73	52	.	.	.
hansen-can-2020	74	1	[	[	-LRB-
hansen-can-2020	74	2	9	9	CD
hansen-can-2020	74	3	:	:	:
hansen-can-2020	74	4	Support	Support	NNP
hansen-can-2020	74	5	Vector	Vector	NNP
hansen-can-2020	74	6	Machines	Machines	NNPS
hansen-can-2020	74	7	(	(	-LRB-
hansen-can-2020	74	8	SVMs	svm	NNS
hansen-can-2020	74	9	)	)	-RRB-
hansen-can-2020	74	10	are	be	VBP
hansen-can-2020	74	11	machine	machine	NN
hansen-can-2020	74	12	learning	learning	NN
hansen-can-2020	74	13	models	model	NNS
hansen-can-2020	74	14	which	which	WDT
hansen-can-2020	74	15	are	be	VBP
hansen-can-2020	74	16	trained	train	VBN
hansen-can-2020	74	17	using	use	VBG
hansen-can-2020	74	18	a	a	DT
hansen-can-2020	74	19	pre	pre	JJ
hansen-can-2020	74	20	-	-	JJ
hansen-can-2020	74	21	classified	classified	JJ
hansen-can-2020	74	22	corpus	corpus	NN
hansen-can-2020	74	23	to	to	TO
hansen-can-2020	74	24	split	split	VB
hansen-can-2020	74	25	a	a	DT
hansen-can-2020	74	26	vector	vector	NN
hansen-can-2020	74	27	space	space	NN
hansen-can-2020	74	28	into	into	IN
hansen-can-2020	74	29	a	a	DT
hansen-can-2020	74	30	set	set	NN
hansen-can-2020	74	31	of	of	IN
hansen-can-2020	74	32	differentiated	differentiate	VBN
hansen-can-2020	74	33	areas	area	NNS
hansen-can-2020	74	34	(	(	-LRB-
hansen-can-2020	74	35	or	or	CC
hansen-can-2020	74	36	categories	category	NNS
hansen-can-2020	74	37	)	)	-RRB-
hansen-can-2020	74	38	and	and	CC
hansen-can-2020	74	39	then	then	RB
hansen-can-2020	74	40	attempt	attempt	VB
hansen-can-2020	74	41	to	to	TO
hansen-can-2020	74	42	classify	classify	VB
hansen-can-2020	74	43	new	new	JJ
hansen-can-2020	74	44	items	item	NNS
hansen-can-2020	74	45	through	through	IN
hansen-can-2020	74	46	where	where	WRB
hansen-can-2020	74	47	in	in	IN
hansen-can-2020	74	48	the	the	DT
hansen-can-2020	74	49	vector	vector	NN
hansen-can-2020	74	50	space	space	NN
hansen-can-2020	74	51	the	the	DT
hansen-can-2020	74	52	trained	train	VBN
hansen-can-2020	74	53	model	model	NN
hansen-can-2020	74	54	places	place	VBZ
hansen-can-2020	74	55	them	-PRON-	PRP
hansen-can-2020	74	56	.	.	.
hansen-can-2020	75	1	For	for	IN
hansen-can-2020	75	2	a	a	DT
hansen-can-2020	75	3	more	more	JJR
hansen-can-2020	75	4	in	in	IN
hansen-can-2020	75	5	-	-	HYPH
hansen-can-2020	75	6	depth	depth	NN
hansen-can-2020	75	7	,	,	,
hansen-can-2020	75	8	technical	technical	JJ
hansen-can-2020	75	9	explanation	explanation	NN
hansen-can-2020	75	10	,	,	,
hansen-can-2020	75	11	see	see	VBP
hansen-can-2020	75	12	:	:	:
hansen-can-2020	75	13	https://towardsdatascience.com/support-vector-machine-introduction-to-machine-learning-algorithms-934a444fca47	https://towardsdatascience.com/support-vector-machine-introduction-to-machine-learning-algorithms-934a444fca47	NNP
hansen-can-2020	75	14	]	]	-RRB-
hansen-can-2020	75	15	[	[	-LRB-
hansen-can-2020	75	16	10	10	CD
hansen-can-2020	75	17	:	:	:
hansen-can-2020	75	18	Augmented	augment	VBN
hansen-can-2020	75	19	term	term	NN
hansen-can-2020	75	20	frequency	frequency	NN
hansen-can-2020	75	21	refers	refer	VBZ
hansen-can-2020	75	22	to	to	IN
hansen-can-2020	75	23	the	the	DT
hansen-can-2020	75	24	number	number	NN
hansen-can-2020	75	25	of	of	IN
hansen-can-2020	75	26	times	time	NNS
hansen-can-2020	75	27	a	a	DT
hansen-can-2020	75	28	term	term	NN
hansen-can-2020	75	29	occurs	occur	VBZ
hansen-can-2020	75	30	in	in	IN
hansen-can-2020	75	31	the	the	DT
hansen-can-2020	75	32	document	document	NN
hansen-can-2020	75	33	divided	divide	VBN
hansen-can-2020	75	34	by	by	IN
hansen-can-2020	75	35	the	the	DT
hansen-can-2020	75	36	number	number	NN
hansen-can-2020	75	37	of	of	IN
hansen-can-2020	75	38	times	time	NNS
hansen-can-2020	75	39	the	the	DT
hansen-can-2020	75	40	most	most	RBS
hansen-can-2020	75	41	frequent	frequent	JJ
hansen-can-2020	75	42	occurring	occurring	NN
hansen-can-2020	75	43	term	term	NN
hansen-can-2020	75	44	appears	appear	VBZ
hansen-can-2020	75	45	in	in	IN
hansen-can-2020	75	46	the	the	DT
hansen-can-2020	75	47	document	document	NN
hansen-can-2020	75	48	.	.	.
hansen-can-2020	75	49	]	]	-RRB-
hansen-can-2020	76	1	[	[	-LRB-
hansen-can-2020	76	2	11	11	CD
hansen-can-2020	76	3	:	:	:
hansen-can-2020	76	4	The	the	DT
hansen-can-2020	76	5	decision	decision	NN
hansen-can-2020	76	6	threshold	threshold	NN
hansen-can-2020	76	7	is	be	VBZ
hansen-can-2020	76	8	the	the	DT
hansen-can-2020	76	9	cut	cut	NN
hansen-can-2020	76	10	-	-	HYPH
hansen-can-2020	76	11	off	off	RP
hansen-can-2020	76	12	for	for	IN
hansen-can-2020	76	13	how	how	WRB
hansen-can-2020	76	14	close	close	RB
hansen-can-2020	76	15	to	to	IN
hansen-can-2020	76	16	a	a	DT
hansen-can-2020	76	17	category	category	NN
hansen-can-2020	76	18	the	the	DT
hansen-can-2020	76	19	SVM	SVM	NNP
hansen-can-2020	76	20	must	must	MD
hansen-can-2020	76	21	determine	determine	VB
hansen-can-2020	76	22	an	an	DT
hansen-can-2020	76	23	item	item	NN
hansen-can-2020	76	24	to	to	TO
hansen-can-2020	76	25	be	be	VB
hansen-can-2020	76	26	in	in	IN
hansen-can-2020	76	27	order	order	NN
hansen-can-2020	76	28	for	for	IN
hansen-can-2020	76	29	it	-PRON-	PRP
hansen-can-2020	76	30	to	to	TO
hansen-can-2020	76	31	be	be	VB
hansen-can-2020	76	32	assigned	assign	VBN
hansen-can-2020	76	33	that	that	DT
hansen-can-2020	76	34	category	category	NN
hansen-can-2020	76	35	.	.	.
hansen-can-2020	77	1	Řehůřek	Řehůřek	NNP
hansen-can-2020	77	2	and	and	CC
hansen-can-2020	77	3	Sojka	Sojka	NNP
hansen-can-2020	77	4	’s	’s	POS
hansen-can-2020	77	5	work	work	NN
hansen-can-2020	77	6	varied	vary	VBD
hansen-can-2020	77	7	this	this	DT
hansen-can-2020	77	8	threshold	threshold	NN
hansen-can-2020	77	9	dynamically	dynamically	RB
hansen-can-2020	77	10	.	.	.
hansen-can-2020	77	11	]	]	-RRB-
hansen-can-2020	78	1	[	[	-LRB-
hansen-can-2020	78	2	12	12	CD
hansen-can-2020	78	3	:	:	:
hansen-can-2020	78	4	Score	score	RB
hansen-can-2020	78	5	-	-	HYPH
hansen-can-2020	78	6	based	base	VBN
hansen-can-2020	78	7	local	local	JJ
hansen-can-2020	78	8	optimization	optimization	NN
hansen-can-2020	78	9	,	,	,
hansen-can-2020	78	10	or	or	CC
hansen-can-2020	78	11	s	s	NN
hansen-can-2020	78	12	-	-	HYPH
hansen-can-2020	78	13	cut	cut	VBN
hansen-can-2020	78	14	,	,	,
hansen-can-2020	78	15	allows	allow	VBZ
hansen-can-2020	78	16	a	a	DT
hansen-can-2020	78	17	machine	machine	NN
hansen-can-2020	78	18	-	-	HYPH
hansen-can-2020	78	19	learning	learn	VBG
hansen-can-2020	78	20	model	model	NN
hansen-can-2020	78	21	to	to	TO
hansen-can-2020	78	22	set	set	VB
hansen-can-2020	78	23	different	different	JJ
hansen-can-2020	78	24	thresholds	threshold	NNS
hansen-can-2020	78	25	for	for	IN
hansen-can-2020	78	26	each	each	DT
hansen-can-2020	78	27	category	category	NN
hansen-can-2020	78	28	with	with	IN
hansen-can-2020	78	29	an	an	DT
hansen-can-2020	78	30	emphasis	emphasis	NN
hansen-can-2020	78	31	on	on	IN
hansen-can-2020	78	32	local	local	JJ
hansen-can-2020	78	33	,	,	,
hansen-can-2020	78	34	or	or	CC
hansen-can-2020	78	35	category	category	NN
hansen-can-2020	78	36	,	,	,
hansen-can-2020	78	37	instead	instead	RB
hansen-can-2020	78	38	of	of	IN
hansen-can-2020	78	39	global	global	JJ
hansen-can-2020	78	40	performance	performance	NN
hansen-can-2020	78	41	.	.	.
hansen-can-2020	78	42	]	]	-RRB-
hansen-can-2020	79	1	[	[	-LRB-
hansen-can-2020	79	2	13	13	CD
hansen-can-2020	79	3	:	:	:
hansen-can-2020	79	4	Term	term	NN
hansen-can-2020	79	5	frequency	frequency	NN
hansen-can-2020	79	6	–	–	:
hansen-can-2020	79	7	inverse	inverse	NN
hansen-can-2020	79	8	document	document	NN
hansen-can-2020	79	9	frequency	frequency	NN
hansen-can-2020	79	10	provides	provide	VBZ
hansen-can-2020	79	11	a	a	DT
hansen-can-2020	79	12	weight	weight	NN
hansen-can-2020	79	13	for	for	IN
hansen-can-2020	79	14	terms	term	NNS
hansen-can-2020	79	15	depending	depend	VBG
hansen-can-2020	79	16	on	on	IN
hansen-can-2020	79	17	how	how	WRB
hansen-can-2020	79	18	frequently	frequently	RB
hansen-can-2020	79	19	it	-PRON-	PRP
hansen-can-2020	79	20	occurs	occur	VBZ
hansen-can-2020	79	21	across	across	IN
hansen-can-2020	79	22	the	the	DT
hansen-can-2020	79	23	corpus	corpus	NN
hansen-can-2020	79	24	.	.	.
hansen-can-2020	80	1	A	a	DT
hansen-can-2020	80	2	term	term	NN
hansen-can-2020	80	3	which	which	WDT
hansen-can-2020	80	4	occurs	occur	VBZ
hansen-can-2020	80	5	rarely	rarely	RB
hansen-can-2020	80	6	across	across	IN
hansen-can-2020	80	7	the	the	DT
hansen-can-2020	80	8	corpus	corpus	NN
hansen-can-2020	80	9	but	but	CC
hansen-can-2020	80	10	with	with	IN
hansen-can-2020	80	11	a	a	DT
hansen-can-2020	80	12	high	high	JJ
hansen-can-2020	80	13	frequency	frequency	NN
hansen-can-2020	80	14	within	within	IN
hansen-can-2020	80	15	a	a	DT
hansen-can-2020	80	16	single	single	JJ
hansen-can-2020	80	17	document	document	NN
hansen-can-2020	80	18	will	will	MD
hansen-can-2020	80	19	have	have	VB
hansen-can-2020	80	20	a	a	DT
hansen-can-2020	80	21	higher	high	JJR
hansen-can-2020	80	22	weight	weight	NN
hansen-can-2020	80	23	when	when	WRB
hansen-can-2020	80	24	classifying	classify	VBG
hansen-can-2020	80	25	the	the	DT
hansen-can-2020	80	26	document	document	NN
hansen-can-2020	80	27	in	in	IN
hansen-can-2020	80	28	question	question	NN
hansen-can-2020	80	29	.	.	.
hansen-can-2020	80	30	]	]	-RRB-
hansen-can-2020	81	1	[	[	-LRB-
hansen-can-2020	81	2	14	14	CD
hansen-can-2020	81	3	:	:	:
hansen-can-2020	81	4	A	a	DT
hansen-can-2020	81	5	Euclidean	euclidean	JJ
hansen-can-2020	81	6	norm	norm	NN
hansen-can-2020	81	7	provides	provide	VBZ
hansen-can-2020	81	8	the	the	DT
hansen-can-2020	81	9	distance	distance	NN
hansen-can-2020	81	10	from	from	IN
hansen-can-2020	81	11	the	the	DT
hansen-can-2020	81	12	origin	origin	NN
hansen-can-2020	81	13	to	to	IN
hansen-can-2020	81	14	a	a	DT
hansen-can-2020	81	15	point	point	NN
hansen-can-2020	81	16	in	in	IN
hansen-can-2020	81	17	an	an	DT
hansen-can-2020	81	18	n	n	JJ
hansen-can-2020	81	19	-	-	HYPH
hansen-can-2020	81	20	dimensional	dimensional	JJ
hansen-can-2020	81	21	space	space	NN
hansen-can-2020	81	22	.	.	.
hansen-can-2020	82	1	It	-PRON-	PRP
hansen-can-2020	82	2	is	be	VBZ
hansen-can-2020	82	3	calculated	calculate	VBN
hansen-can-2020	82	4	by	by	IN
hansen-can-2020	82	5	taking	take	VBG
hansen-can-2020	82	6	the	the	DT
hansen-can-2020	82	7	square	square	JJ
hansen-can-2020	82	8	root	root	NN
hansen-can-2020	82	9	of	of	IN
hansen-can-2020	82	10	the	the	DT
hansen-can-2020	82	11	sum	sum	NN
hansen-can-2020	82	12	of	of	IN
hansen-can-2020	82	13	the	the	DT
hansen-can-2020	82	14	squares	square	NNS
hansen-can-2020	82	15	of	of	IN
hansen-can-2020	82	16	all	all	DT
hansen-can-2020	82	17	coordinate	coordinate	JJ
hansen-can-2020	82	18	values	value	NNS
hansen-can-2020	82	19	.	.	.
hansen-can-2020	82	20	]	]	-RRB-
hansen-can-2020	83	1	Now	now	RB
hansen-can-2020	83	2	let	let	VB
hansen-can-2020	83	3	us	-PRON-	PRP
hansen-can-2020	83	4	shift	shift	VB
hansen-can-2020	83	5	from	from	IN
hansen-can-2020	83	6	mathematics	mathematics	NN
hansen-can-2020	83	7	-	-	HYPH
hansen-can-2020	83	8	specific	specific	JJ
hansen-can-2020	83	9	categorization	categorization	NN
hansen-can-2020	83	10	to	to	TO
hansen-can-2020	83	11	subject	subject	JJ
hansen-can-2020	83	12	categorization	categorization	NN
hansen-can-2020	83	13	in	in	IN
hansen-can-2020	83	14	general	general	JJ
hansen-can-2020	83	15	and	and	CC
hansen-can-2020	83	16	look	look	VB
hansen-can-2020	83	17	at	at	IN
hansen-can-2020	83	18	the	the	DT
hansen-can-2020	83	19	work	work	NN
hansen-can-2020	83	20	Microsoft	Microsoft	NNP
hansen-can-2020	83	21	has	have	VBZ
hansen-can-2020	83	22	done	do	VBN
hansen-can-2020	83	23	assigning	assigning	NN
hansen-can-2020	83	24	Fields	Fields	NNP
hansen-can-2020	83	25	of	of	IN
hansen-can-2020	83	26	Study	Study	NNP
hansen-can-2020	83	27	(	(	-LRB-
hansen-can-2020	83	28	FoS	FoS	NNP
hansen-can-2020	83	29	)	)	-RRB-
hansen-can-2020	83	30	in	in	IN
hansen-can-2020	83	31	the	the	DT
hansen-can-2020	83	32	Microsoft	Microsoft	NNP
hansen-can-2020	83	33	Academic	Academic	NNP
hansen-can-2020	83	34	Graph	Graph	NNP
hansen-can-2020	83	35	(	(	-LRB-
hansen-can-2020	83	36	MAG	MAG	NNP
hansen-can-2020	83	37	)	)	-RRB-
hansen-can-2020	83	38	which	which	WDT
hansen-can-2020	83	39	is	be	VBZ
hansen-can-2020	83	40	used	use	VBN
hansen-can-2020	83	41	to	to	TO
hansen-can-2020	83	42	create	create	VB
hansen-can-2020	83	43	their	-PRON-	PRP$
hansen-can-2020	83	44	Microsoft	Microsoft	NNP
hansen-can-2020	83	45	Academic	Academic	NNP
hansen-can-2020	83	46	article	article	NN
hansen-can-2020	83	47	search	search	NN
hansen-can-2020	83	48	product	product	NN
hansen-can-2020	83	49	.	.	.
hansen-can-2020	84	1	[	[	-LRB-
hansen-can-2020	84	2	footnoteRef:15	footnoteref:15	CD
hansen-can-2020	84	3	]	]	-RRB-
hansen-can-2020	84	4	While	while	IN
hansen-can-2020	84	5	the	the	DT
hansen-can-2020	84	6	MAG	MAG	NNP
hansen-can-2020	84	7	FoS	fos	NN
hansen-can-2020	84	8	project	project	NN
hansen-can-2020	84	9	is	be	VBZ
hansen-can-2020	84	10	also	also	RB
hansen-can-2020	84	11	attempting	attempt	VBG
hansen-can-2020	84	12	to	to	TO
hansen-can-2020	84	13	categorize	categorize	VB
hansen-can-2020	84	14	articles	article	NNS
hansen-can-2020	84	15	for	for	IN
hansen-can-2020	84	16	proper	proper	JJ
hansen-can-2020	84	17	indexing	indexing	NN
hansen-can-2020	84	18	and	and	CC
hansen-can-2020	84	19	search	search	NN
hansen-can-2020	84	20	,	,	,
hansen-can-2020	84	21	it	-PRON-	PRP
hansen-can-2020	84	22	represents	represent	VBZ
hansen-can-2020	84	23	the	the	DT
hansen-can-2020	84	24	second	second	JJ
hansen-can-2020	84	25	path	path	NN
hansen-can-2020	84	26	which	which	WDT
hansen-can-2020	84	27	is	be	VBZ
hansen-can-2020	84	28	taken	take	VBN
hansen-can-2020	84	29	by	by	IN
hansen-can-2020	84	30	automated	automate	VBN
hansen-can-2020	84	31	categorization	categorization	NN
hansen-can-2020	84	32	projects	project	NNS
hansen-can-2020	84	33	:	:	:
hansen-can-2020	84	34	using	use	VBG
hansen-can-2020	84	35	machine	machine	NN
hansen-can-2020	84	36	learning	learning	NN
hansen-can-2020	84	37	techniques	technique	NNS
hansen-can-2020	84	38	to	to	TO
hansen-can-2020	84	39	both	both	DT
hansen-can-2020	84	40	create	create	VB
hansen-can-2020	84	41	the	the	DT
hansen-can-2020	84	42	taxonomy	taxonomy	NN
hansen-can-2020	84	43	and	and	CC
hansen-can-2020	84	44	to	to	TO
hansen-can-2020	84	45	classify	classify	VB
hansen-can-2020	84	46	.	.	.
hansen-can-2020	85	1	[	[	-LRB-
hansen-can-2020	85	2	15	15	CD
hansen-can-2020	85	3	:	:	:
hansen-can-2020	85	4	https://academic.microsoft.com/	https://academic.microsoft.com/	LS
hansen-can-2020	85	5	]	]	-RRB-
hansen-can-2020	85	6	Microsoft	Microsoft	NNP
hansen-can-2020	85	7	took	take	VBD
hansen-can-2020	85	8	a	a	DT
hansen-can-2020	85	9	unique	unique	JJ
hansen-can-2020	85	10	approach	approach	NN
hansen-can-2020	85	11	in	in	IN
hansen-can-2020	85	12	the	the	DT
hansen-can-2020	85	13	development	development	NN
hansen-can-2020	85	14	of	of	IN
hansen-can-2020	85	15	their	-PRON-	PRP$
hansen-can-2020	85	16	taxonomy	taxonomy	NN
hansen-can-2020	85	17	.	.	.
hansen-can-2020	86	1	Instead	instead	RB
hansen-can-2020	86	2	of	of	IN
hansen-can-2020	86	3	relying	rely	VBG
hansen-can-2020	86	4	on	on	IN
hansen-can-2020	86	5	the	the	DT
hansen-can-2020	86	6	corpus	corpus	NN
hansen-can-2020	86	7	of	of	IN
hansen-can-2020	86	8	articles	article	NNS
hansen-can-2020	86	9	in	in	IN
hansen-can-2020	86	10	the	the	DT
hansen-can-2020	86	11	MAG	MAG	NNP
hansen-can-2020	86	12	to	to	TO
hansen-can-2020	86	13	develop	develop	VB
hansen-can-2020	86	14	it	-PRON-	PRP
hansen-can-2020	86	15	,	,	,
hansen-can-2020	86	16	they	-PRON-	PRP
hansen-can-2020	86	17	relied	rely	VBD
hansen-can-2020	86	18	primarily	primarily	RB
hansen-can-2020	86	19	on	on	IN
hansen-can-2020	86	20	Wikipedia	Wikipedia	NNP
hansen-can-2020	86	21	for	for	IN
hansen-can-2020	86	22	its	-PRON-	PRP$
hansen-can-2020	86	23	creation	creation	NN
hansen-can-2020	86	24	.	.	.
hansen-can-2020	87	1	They	-PRON-	PRP
hansen-can-2020	87	2	generated	generate	VBD
hansen-can-2020	87	3	an	an	DT
hansen-can-2020	87	4	initial	initial	JJ
hansen-can-2020	87	5	seed	seed	NN
hansen-can-2020	87	6	by	by	IN
hansen-can-2020	87	7	referencing	reference	VBG
hansen-can-2020	87	8	the	the	DT
hansen-can-2020	87	9	Science	Science	NNP
hansen-can-2020	87	10	Metrix	Metrix	NNP
hansen-can-2020	87	11	classification	classification	NN
hansen-can-2020	87	12	scheme[footnoteRef:16	scheme[footnoteref:16	NN
hansen-can-2020	87	13	]	]	-RRB-
hansen-can-2020	87	14	and	and	CC
hansen-can-2020	87	15	a	a	DT
hansen-can-2020	87	16	couple	couple	NN
hansen-can-2020	87	17	thousand	thousand	CD
hansen-can-2020	87	18	FoS	FoS	NNP
hansen-can-2020	87	19	Wikipedia	Wikipedia	NNP
hansen-can-2020	87	20	articles	article	NNS
hansen-can-2020	87	21	they	-PRON-	PRP
hansen-can-2020	87	22	identified	identify	VBD
hansen-can-2020	87	23	internally	internally	RB
hansen-can-2020	87	24	.	.	.
hansen-can-2020	88	1	They	-PRON-	PRP
hansen-can-2020	88	2	then	then	RB
hansen-can-2020	88	3	used	use	VBD
hansen-can-2020	88	4	an	an	DT
hansen-can-2020	88	5	iterative	iterative	JJ
hansen-can-2020	88	6	process	process	NN
hansen-can-2020	88	7	to	to	TO
hansen-can-2020	88	8	identify	identify	VB
hansen-can-2020	88	9	more	more	JJR
hansen-can-2020	88	10	FoS	fos	NN
hansen-can-2020	88	11	in	in	IN
hansen-can-2020	88	12	Wikipedia	Wikipedia	NNP
hansen-can-2020	88	13	based	base	VBN
hansen-can-2020	88	14	on	on	IN
hansen-can-2020	88	15	whether	whether	IN
hansen-can-2020	88	16	they	-PRON-	PRP
hansen-can-2020	88	17	were	be	VBD
hansen-can-2020	88	18	linked	link	VBN
hansen-can-2020	88	19	to	to	IN
hansen-can-2020	88	20	Wikipedia	Wikipedia	NNP
hansen-can-2020	88	21	articles	article	NNS
hansen-can-2020	88	22	that	that	WDT
hansen-can-2020	88	23	were	be	VBD
hansen-can-2020	88	24	already	already	RB
hansen-can-2020	88	25	identified	identify	VBN
hansen-can-2020	88	26	as	as	IN
hansen-can-2020	88	27	FoS	fos	NN
hansen-can-2020	88	28	and	and	CC
hansen-can-2020	88	29	whether	whether	IN
hansen-can-2020	88	30	the	the	DT
hansen-can-2020	88	31	new	new	JJ
hansen-can-2020	88	32	articles	article	NNS
hansen-can-2020	88	33	represented	represent	VBD
hansen-can-2020	88	34	valid	valid	JJ
hansen-can-2020	88	35	entity	entity	NN
hansen-can-2020	88	36	types	type	NNS
hansen-can-2020	88	37	–	–	:
hansen-can-2020	88	38	e.g.	e.g.	RB
hansen-can-2020	89	1	an	an	DT
hansen-can-2020	89	2	entity	entity	NN
hansen-can-2020	89	3	type	type	NN
hansen-can-2020	89	4	of	of	IN
hansen-can-2020	89	5	protein	protein	NN
hansen-can-2020	89	6	would	would	MD
hansen-can-2020	89	7	be	be	VB
hansen-can-2020	89	8	added	add	VBN
hansen-can-2020	89	9	and	and	CC
hansen-can-2020	89	10	an	an	DT
hansen-can-2020	89	11	entity	entity	NN
hansen-can-2020	89	12	type	type	NN
hansen-can-2020	89	13	of	of	IN
hansen-can-2020	89	14	person	person	NN
hansen-can-2020	89	15	would	would	MD
hansen-can-2020	89	16	be	be	VB
hansen-can-2020	89	17	excluded	exclude	VBN
hansen-can-2020	89	18	(	(	-LRB-
hansen-can-2020	89	19	Shen	Shen	NNP
hansen-can-2020	89	20	,	,	,
hansen-can-2020	89	21	Ma	Ma	NNP
hansen-can-2020	89	22	,	,	,
hansen-can-2020	89	23	and	and	CC
hansen-can-2020	89	24	Wang	Wang	NNP
hansen-can-2020	89	25	2018	2018	CD
hansen-can-2020	89	26	,	,	,
hansen-can-2020	89	27	3	3	CD
hansen-can-2020	89	28	)	)	-RRB-
hansen-can-2020	89	29	.	.	.
hansen-can-2020	90	1	This	this	DT
hansen-can-2020	90	2	work	work	NN
hansen-can-2020	90	3	allowed	allow	VBD
hansen-can-2020	90	4	Microsoft	Microsoft	NNP
hansen-can-2020	90	5	to	to	TO
hansen-can-2020	90	6	develop	develop	VB
hansen-can-2020	90	7	a	a	DT
hansen-can-2020	90	8	list	list	NN
hansen-can-2020	90	9	of	of	IN
hansen-can-2020	90	10	more	more	JJR
hansen-can-2020	90	11	than	than	IN
hansen-can-2020	90	12	200,000	200,000	CD
hansen-can-2020	90	13	Fields	field	NNS
hansen-can-2020	90	14	of	of	IN
hansen-can-2020	90	15	Study	Study	NNP
hansen-can-2020	90	16	for	for	IN
hansen-can-2020	90	17	use	use	NN
hansen-can-2020	90	18	as	as	IN
hansen-can-2020	90	19	categories	category	NNS
hansen-can-2020	90	20	in	in	IN
hansen-can-2020	90	21	the	the	DT
hansen-can-2020	90	22	MAG	MAG	NNP
hansen-can-2020	90	23	.	.	.
hansen-can-2020	91	1	[	[	-LRB-
hansen-can-2020	91	2	16	16	CD
hansen-can-2020	91	3	:	:	:
hansen-can-2020	91	4	http://science-metrix.com/?q=en/classification	http://science-metrix.com/?q=en/classification	NNP
hansen-can-2020	91	5	]	]	-RRB-
hansen-can-2020	91	6	Microsoft	Microsoft	NNP
hansen-can-2020	91	7	then	then	RB
hansen-can-2020	91	8	used	use	VBD
hansen-can-2020	91	9	machine	machine	NN
hansen-can-2020	91	10	learning	learning	NN
hansen-can-2020	91	11	techniques	technique	NNS
hansen-can-2020	91	12	to	to	TO
hansen-can-2020	91	13	apply	apply	VB
hansen-can-2020	91	14	these	these	DT
hansen-can-2020	91	15	FoS	fos	NN
hansen-can-2020	91	16	to	to	IN
hansen-can-2020	91	17	their	-PRON-	PRP$
hansen-can-2020	91	18	corpus	corpus	NN
hansen-can-2020	91	19	of	of	IN
hansen-can-2020	91	20	over	over	IN
hansen-can-2020	91	21	140	140	CD
hansen-can-2020	91	22	million	million	CD
hansen-can-2020	91	23	academic	academic	JJ
hansen-can-2020	91	24	articles	article	NNS
hansen-can-2020	91	25	.	.	.
hansen-can-2020	92	1	The	the	DT
hansen-can-2020	92	2	specific	specific	JJ
hansen-can-2020	92	3	techniques	technique	NNS
hansen-can-2020	92	4	are	be	VBP
hansen-can-2020	92	5	not	not	RB
hansen-can-2020	92	6	as	as	RB
hansen-can-2020	92	7	clear	clear	JJ
hansen-can-2020	92	8	as	as	IN
hansen-can-2020	92	9	they	-PRON-	PRP
hansen-can-2020	92	10	were	be	VBD
hansen-can-2020	92	11	with	with	IN
hansen-can-2020	92	12	the	the	DT
hansen-can-2020	92	13	previous	previous	JJ
hansen-can-2020	92	14	examples	example	NNS
hansen-can-2020	92	15	,	,	,
hansen-can-2020	92	16	likely	likely	RB
hansen-can-2020	92	17	due	due	IN
hansen-can-2020	92	18	to	to	IN
hansen-can-2020	92	19	Microsoft	Microsoft	NNP
hansen-can-2020	92	20	protecting	protect	VBG
hansen-can-2020	92	21	their	-PRON-	PRP$
hansen-can-2020	92	22	specific	specific	JJ
hansen-can-2020	92	23	methods	method	NNS
hansen-can-2020	92	24	from	from	IN
hansen-can-2020	92	25	competitors	competitor	NNS
hansen-can-2020	92	26	,	,	,
hansen-can-2020	92	27	but	but	CC
hansen-can-2020	92	28	the	the	DT
hansen-can-2020	92	29	article	article	NN
hansen-can-2020	92	30	published	publish	VBN
hansen-can-2020	92	31	to	to	IN
hansen-can-2020	92	32	the	the	DT
hansen-can-2020	92	33	arXiv	arXiv	NNP
hansen-can-2020	92	34	by	by	IN
hansen-can-2020	92	35	their	-PRON-	PRP$
hansen-can-2020	92	36	researchers	researcher	NNS
hansen-can-2020	92	37	(	(	-LRB-
hansen-can-2020	92	38	Shen	Shen	NNP
hansen-can-2020	92	39	,	,	,
hansen-can-2020	92	40	Ma	Ma	NNP
hansen-can-2020	92	41	,	,	,
hansen-can-2020	92	42	and	and	CC
hansen-can-2020	92	43	Wang	Wang	NNP
hansen-can-2020	92	44	2018	2018	CD
hansen-can-2020	92	45	)	)	-RRB-
hansen-can-2020	92	46	and	and	CC
hansen-can-2020	92	47	the	the	DT
hansen-can-2020	92	48	write	write	NN
hansen-can-2020	92	49	up	up	RP
hansen-can-2020	92	50	on	on	IN
hansen-can-2020	92	51	the	the	DT
hansen-can-2020	92	52	MAG	MAG	NNP
hansen-can-2020	92	53	website	website	NN
hansen-can-2020	92	54	does	do	VBZ
hansen-can-2020	92	55	make	make	VB
hansen-can-2020	92	56	it	-PRON-	PRP
hansen-can-2020	92	57	clear	clear	JJ
hansen-can-2020	92	58	they	-PRON-	PRP
hansen-can-2020	92	59	used	use	VBD
hansen-can-2020	92	60	vector	vector	NN
hansen-can-2020	92	61	based	base	VBN
hansen-can-2020	92	62	convolutional	convolutional	JJ
hansen-can-2020	92	63	neural	neural	JJ
hansen-can-2020	92	64	networks	network	NNS
hansen-can-2020	92	65	which	which	WDT
hansen-can-2020	92	66	relied	rely	VBD
hansen-can-2020	92	67	on	on	IN
hansen-can-2020	92	68	Skip	Skip	NNP
hansen-can-2020	92	69	-	-	HYPH
hansen-can-2020	92	70	gram	gram	NNP
hansen-can-2020	92	71	(	(	-LRB-
hansen-can-2020	92	72	Mikolov	Mikolov	NNP
hansen-can-2020	92	73	et	et	FW
hansen-can-2020	92	74	al	al	NNP
hansen-can-2020	92	75	.	.	.
hansen-can-2020	93	1	2013	2013	LS
hansen-can-2020	93	2	)	)	-RRB-
hansen-can-2020	93	3	embeddings	embedding	NNS
hansen-can-2020	93	4	and	and	CC
hansen-can-2020	93	5	bag	bag	NN
hansen-can-2020	93	6	-	-	HYPH
hansen-can-2020	93	7	of	of	IN
hansen-can-2020	93	8	-	-	HYPH
hansen-can-2020	93	9	words	word	NNS
hansen-can-2020	93	10	/	/	SYM
hansen-can-2020	93	11	entities	entity	NNS
hansen-can-2020	93	12	features	feature	VBZ
hansen-can-2020	93	13	to	to	TO
hansen-can-2020	93	14	create	create	VB
hansen-can-2020	93	15	their	-PRON-	PRP$
hansen-can-2020	93	16	vectors	vector	NNS
hansen-can-2020	93	17	(	(	-LRB-
hansen-can-2020	93	18	“	"	``
hansen-can-2020	93	19	Microsoft	Microsoft	NNP
hansen-can-2020	93	20	Academic	Academic	NNP
hansen-can-2020	93	21	Increases	Increases	NNPS
hansen-can-2020	93	22	Power	Power	NNP
hansen-can-2020	93	23	of	of	IN
hansen-can-2020	93	24	Semantic	Semantic	NNP
hansen-can-2020	93	25	Search	Search	NNP
hansen-can-2020	93	26	by	by	IN
hansen-can-2020	93	27	Adding	add	VBG
hansen-can-2020	93	28	More	More	JJR
hansen-can-2020	93	29	Fields	field	NNS
hansen-can-2020	93	30	of	of	IN
hansen-can-2020	93	31	Study	Study	NNP
hansen-can-2020	93	32	-	-	HYPH
hansen-can-2020	93	33	Microsoft	Microsoft	NNP
hansen-can-2020	93	34	Research	Research	NNP
hansen-can-2020	93	35	”	"	''
hansen-can-2020	93	36	2018	2018	CD
hansen-can-2020	93	37	)	)	-RRB-
hansen-can-2020	93	38	.	.	.
hansen-can-2020	94	1	One	one	CD
hansen-can-2020	94	2	really	really	RB
hansen-can-2020	94	3	interesting	interesting	JJ
hansen-can-2020	94	4	part	part	NN
hansen-can-2020	94	5	of	of	IN
hansen-can-2020	94	6	the	the	DT
hansen-can-2020	94	7	machine	machine	NN
hansen-can-2020	94	8	learning	learning	NN
hansen-can-2020	94	9	method	method	NN
hansen-can-2020	94	10	used	use	VBN
hansen-can-2020	94	11	by	by	IN
hansen-can-2020	94	12	Microsoft	Microsoft	NNP
hansen-can-2020	94	13	was	be	VBD
hansen-can-2020	94	14	that	that	IN
hansen-can-2020	94	15	it	-PRON-	PRP
hansen-can-2020	94	16	did	do	VBD
hansen-can-2020	94	17	not	not	RB
hansen-can-2020	94	18	rely	rely	VB
hansen-can-2020	94	19	only	only	RB
hansen-can-2020	94	20	on	on	IN
hansen-can-2020	94	21	information	information	NN
hansen-can-2020	94	22	from	from	IN
hansen-can-2020	94	23	the	the	DT
hansen-can-2020	94	24	article	article	NN
hansen-can-2020	94	25	being	be	VBG
hansen-can-2020	94	26	categorized	categorize	VBN
hansen-can-2020	94	27	.	.	.
hansen-can-2020	95	1	It	-PRON-	PRP
hansen-can-2020	95	2	also	also	RB
hansen-can-2020	95	3	utilized	utilize	VBD
hansen-can-2020	95	4	the	the	DT
hansen-can-2020	95	5	citations	citation	NNS
hansen-can-2020	95	6	to	to	IN
hansen-can-2020	95	7	and	and	CC
hansen-can-2020	95	8	references	reference	NNS
hansen-can-2020	95	9	from	from	IN
hansen-can-2020	95	10	information	information	NN
hansen-can-2020	95	11	about	about	IN
hansen-can-2020	95	12	the	the	DT
hansen-can-2020	95	13	article	article	NN
hansen-can-2020	95	14	in	in	IN
hansen-can-2020	95	15	the	the	DT
hansen-can-2020	95	16	MAG	MAG	NNP
hansen-can-2020	95	17	,	,	,
hansen-can-2020	95	18	and	and	CC
hansen-can-2020	95	19	used	use	VBD
hansen-can-2020	95	20	the	the	DT
hansen-can-2020	95	21	FoS	fos	NN
hansen-can-2020	95	22	the	the	DT
hansen-can-2020	95	23	citations	citation	NNS
hansen-can-2020	95	24	and	and	CC
hansen-can-2020	95	25	references	reference	NNS
hansen-can-2020	95	26	were	be	VBD
hansen-can-2020	95	27	assigned	assign	VBN
hansen-can-2020	95	28	in	in	IN
hansen-can-2020	95	29	order	order	NN
hansen-can-2020	95	30	to	to	TO
hansen-can-2020	95	31	influence	influence	VB
hansen-can-2020	95	32	the	the	DT
hansen-can-2020	95	33	FoS	fos	NN
hansen-can-2020	95	34	of	of	IN
hansen-can-2020	95	35	the	the	DT
hansen-can-2020	95	36	original	original	JJ
hansen-can-2020	95	37	article	article	NN
hansen-can-2020	95	38	.	.	.
hansen-can-2020	96	1	The	the	DT
hansen-can-2020	96	2	identification	identification	NN
hansen-can-2020	96	3	of	of	IN
hansen-can-2020	96	4	potential	potential	JJ
hansen-can-2020	96	5	FoS	fos	NN
hansen-can-2020	96	6	and	and	CC
hansen-can-2020	96	7	their	-PRON-	PRP$
hansen-can-2020	96	8	assignment	assignment	NN
hansen-can-2020	96	9	to	to	IN
hansen-can-2020	96	10	articles	article	NNS
hansen-can-2020	96	11	was	be	VBD
hansen-can-2020	96	12	only	only	RB
hansen-can-2020	96	13	a	a	DT
hansen-can-2020	96	14	part	part	NN
hansen-can-2020	96	15	of	of	IN
hansen-can-2020	96	16	Microsoft	Microsoft	NNP
hansen-can-2020	96	17	's	's	POS
hansen-can-2020	96	18	purpose	purpose	NN
hansen-can-2020	96	19	.	.	.
hansen-can-2020	97	1	In	in	IN
hansen-can-2020	97	2	order	order	NN
hansen-can-2020	97	3	to	to	TO
hansen-can-2020	97	4	fully	fully	RB
hansen-can-2020	97	5	index	index	VB
hansen-can-2020	97	6	the	the	DT
hansen-can-2020	97	7	MAG	MAG	NNP
hansen-can-2020	97	8	and	and	CC
hansen-can-2020	97	9	make	make	VB
hansen-can-2020	97	10	it	-PRON-	PRP
hansen-can-2020	97	11	searchable	searchable	JJ
hansen-can-2020	97	12	they	-PRON-	PRP
hansen-can-2020	97	13	also	also	RB
hansen-can-2020	97	14	wished	wish	VBD
hansen-can-2020	97	15	to	to	TO
hansen-can-2020	97	16	determine	determine	VB
hansen-can-2020	97	17	the	the	DT
hansen-can-2020	97	18	relationships	relationship	NNS
hansen-can-2020	97	19	between	between	IN
hansen-can-2020	97	20	the	the	DT
hansen-can-2020	97	21	FoS	fos	NN
hansen-can-2020	97	22	;	;	:
hansen-can-2020	97	23	in	in	IN
hansen-can-2020	97	24	other	other	JJ
hansen-can-2020	97	25	words	word	NNS
hansen-can-2020	97	26	they	-PRON-	PRP
hansen-can-2020	97	27	wanted	want	VBD
hansen-can-2020	97	28	to	to	TO
hansen-can-2020	97	29	build	build	VB
hansen-can-2020	97	30	a	a	DT
hansen-can-2020	97	31	hierarchical	hierarchical	JJ
hansen-can-2020	97	32	taxonomy	taxonomy	NN
hansen-can-2020	97	33	.	.	.
hansen-can-2020	98	1	To	to	TO
hansen-can-2020	98	2	achieve	achieve	VB
hansen-can-2020	98	3	this	this	DT
hansen-can-2020	98	4	they	-PRON-	PRP
hansen-can-2020	98	5	used	use	VBD
hansen-can-2020	98	6	the	the	DT
hansen-can-2020	98	7	article	article	NN
hansen-can-2020	98	8	categorizations	categorization	NNS
hansen-can-2020	98	9	and	and	CC
hansen-can-2020	98	10	defined	define	VBD
hansen-can-2020	98	11	a	a	DT
hansen-can-2020	98	12	Field	Field	NNP
hansen-can-2020	98	13	of	of	IN
hansen-can-2020	98	14	Study	Study	NNP
hansen-can-2020	98	15	A	A	NNP
hansen-can-2020	98	16	as	as	IN
hansen-can-2020	98	17	the	the	DT
hansen-can-2020	98	18	parent	parent	NN
hansen-can-2020	98	19	of	of	IN
hansen-can-2020	98	20	B	b	NN
hansen-can-2020	98	21	if	if	IN
hansen-can-2020	98	22	the	the	DT
hansen-can-2020	98	23	articles	article	NNS
hansen-can-2020	98	24	categorized	categorize	VBN
hansen-can-2020	98	25	as	as	IN
hansen-can-2020	98	26	B	B	NNP
hansen-can-2020	98	27	were	be	VBD
hansen-can-2020	98	28	close	close	JJ
hansen-can-2020	98	29	to	to	IN
hansen-can-2020	98	30	a	a	DT
hansen-can-2020	98	31	subset	subset	NN
hansen-can-2020	98	32	of	of	IN
hansen-can-2020	98	33	the	the	DT
hansen-can-2020	98	34	articles	article	NNS
hansen-can-2020	98	35	categorized	categorize	VBN
hansen-can-2020	98	36	as	as	IN
hansen-can-2020	98	37	A	a	DT
hansen-can-2020	98	38	(	(	-LRB-
hansen-can-2020	98	39	a	a	DT
hansen-can-2020	98	40	more	more	RBR
hansen-can-2020	98	41	formal	formal	JJ
hansen-can-2020	98	42	definition	definition	NN
hansen-can-2020	98	43	can	can	MD
hansen-can-2020	98	44	be	be	VB
hansen-can-2020	98	45	found	find	VBN
hansen-can-2020	98	46	in	in	IN
hansen-can-2020	98	47	Shen	Shen	NNP
hansen-can-2020	98	48	,	,	,
hansen-can-2020	98	49	Ma	Ma	NNP
hansen-can-2020	98	50	,	,	,
hansen-can-2020	98	51	and	and	CC
hansen-can-2020	98	52	Wang	Wang	NNP
hansen-can-2020	98	53	2018	2018	CD
hansen-can-2020	98	54	,	,	,
hansen-can-2020	98	55	4	4	CD
hansen-can-2020	98	56	)	)	-RRB-
hansen-can-2020	98	57	.	.	.
hansen-can-2020	99	1	This	this	DT
hansen-can-2020	99	2	work	work	NN
hansen-can-2020	99	3	,	,	,
hansen-can-2020	99	4	which	which	WDT
hansen-can-2020	99	5	created	create	VBD
hansen-can-2020	99	6	a	a	DT
hansen-can-2020	99	7	six	six	CD
hansen-can-2020	99	8	-	-	HYPH
hansen-can-2020	99	9	level	level	NN
hansen-can-2020	99	10	hierarchy	hierarchy	NN
hansen-can-2020	99	11	,	,	,
hansen-can-2020	99	12	was	be	VBD
hansen-can-2020	99	13	mostly	mostly	RB
hansen-can-2020	99	14	automated	automate	VBN
hansen-can-2020	99	15	,	,	,
hansen-can-2020	99	16	but	but	CC
hansen-can-2020	99	17	Microsoft	Microsoft	NNP
hansen-can-2020	99	18	did	do	VBD
hansen-can-2020	99	19	inspect	inspect	VB
hansen-can-2020	99	20	and	and	CC
hansen-can-2020	99	21	manually	manually	RB
hansen-can-2020	99	22	adjust	adjust	VB
hansen-can-2020	99	23	the	the	DT
hansen-can-2020	99	24	relationships	relationship	NNS
hansen-can-2020	99	25	between	between	IN
hansen-can-2020	99	26	FoS	FoS	NNP
hansen-can-2020	99	27	on	on	IN
hansen-can-2020	99	28	the	the	DT
hansen-can-2020	99	29	highest	high	JJS
hansen-can-2020	99	30	two	two	CD
hansen-can-2020	99	31	levels	level	NNS
hansen-can-2020	99	32	.	.	.
hansen-can-2020	100	1	To	to	TO
hansen-can-2020	100	2	evaluate	evaluate	VB
hansen-can-2020	100	3	the	the	DT
hansen-can-2020	100	4	quality	quality	NN
hansen-can-2020	100	5	of	of	IN
hansen-can-2020	100	6	their	-PRON-	PRP$
hansen-can-2020	100	7	FoS	FoS	NNP
hansen-can-2020	100	8	taxonomy	taxonomy	NN
hansen-can-2020	100	9	and	and	CC
hansen-can-2020	100	10	categorization	categorization	NN
hansen-can-2020	100	11	work	work	NN
hansen-can-2020	100	12	,	,	,
hansen-can-2020	100	13	Microsoft	Microsoft	NNP
hansen-can-2020	100	14	randomly	randomly	RB
hansen-can-2020	100	15	sampled	sample	VBD
hansen-can-2020	100	16	data	datum	NNS
hansen-can-2020	100	17	at	at	IN
hansen-can-2020	100	18	each	each	DT
hansen-can-2020	100	19	of	of	IN
hansen-can-2020	100	20	the	the	DT
hansen-can-2020	100	21	three	three	CD
hansen-can-2020	100	22	steps	step	NNS
hansen-can-2020	100	23	of	of	IN
hansen-can-2020	100	24	the	the	DT
hansen-can-2020	100	25	project	project	NN
hansen-can-2020	100	26	and	and	CC
hansen-can-2020	100	27	used	use	VBD
hansen-can-2020	100	28	human	human	JJ
hansen-can-2020	100	29	judges	judge	NNS
hansen-can-2020	100	30	to	to	TO
hansen-can-2020	100	31	assess	assess	VB
hansen-can-2020	100	32	their	-PRON-	PRP$
hansen-can-2020	100	33	accuracy	accuracy	NN
hansen-can-2020	100	34	.	.	.
hansen-can-2020	101	1	The	the	DT
hansen-can-2020	101	2	accuracy	accuracy	NN
hansen-can-2020	101	3	assessments	assessment	NNS
hansen-can-2020	101	4	of	of	IN
hansen-can-2020	101	5	the	the	DT
hansen-can-2020	101	6	three	three	CD
hansen-can-2020	101	7	steps	step	NNS
hansen-can-2020	101	8	were	be	VBD
hansen-can-2020	101	9	not	not	RB
hansen-can-2020	101	10	as	as	RB
hansen-can-2020	101	11	complete	complete	JJ
hansen-can-2020	101	12	as	as	IN
hansen-can-2020	101	13	they	-PRON-	PRP
hansen-can-2020	101	14	would	would	MD
hansen-can-2020	101	15	be	be	VB
hansen-can-2020	101	16	with	with	IN
hansen-can-2020	101	17	the	the	DT
hansen-can-2020	101	18	mathematics	mathematic	NNS
hansen-can-2020	101	19	categorization	categorization	NN
hansen-can-2020	101	20	,	,	,
hansen-can-2020	101	21	as	as	IN
hansen-can-2020	101	22	that	that	DT
hansen-can-2020	101	23	approach	approach	NN
hansen-can-2020	101	24	would	would	MD
hansen-can-2020	101	25	evaluate	evaluate	VB
hansen-can-2020	101	26	terms	term	NNS
hansen-can-2020	101	27	across	across	IN
hansen-can-2020	101	28	the	the	DT
hansen-can-2020	101	29	whole	whole	NN
hansen-can-2020	101	30	of	of	IN
hansen-can-2020	101	31	their	-PRON-	PRP$
hansen-can-2020	101	32	data	data	NN
hansen-can-2020	101	33	sets	set	NNS
hansen-can-2020	101	34	,	,	,
hansen-can-2020	101	35	but	but	CC
hansen-can-2020	101	36	the	the	DT
hansen-can-2020	101	37	projects	project	NNS
hansen-can-2020	101	38	are	be	VBP
hansen-can-2020	101	39	of	of	IN
hansen-can-2020	101	40	very	very	RB
hansen-can-2020	101	41	different	different	JJ
hansen-can-2020	101	42	scales	scale	NNS
hansen-can-2020	101	43	so	so	RB
hansen-can-2020	101	44	different	different	JJ
hansen-can-2020	101	45	methods	method	NNS
hansen-can-2020	101	46	are	be	VBP
hansen-can-2020	101	47	appropriate	appropriate	JJ
hansen-can-2020	101	48	.	.	.
hansen-can-2020	102	1	In	in	IN
hansen-can-2020	102	2	the	the	DT
hansen-can-2020	102	3	end	end	NN
hansen-can-2020	102	4	Microsoft	Microsoft	NNP
hansen-can-2020	102	5	estimates	estimate	VBZ
hansen-can-2020	102	6	the	the	DT
hansen-can-2020	102	7	accuracy	accuracy	NN
hansen-can-2020	102	8	of	of	IN
hansen-can-2020	102	9	the	the	DT
hansen-can-2020	102	10	FoS	fos	NN
hansen-can-2020	102	11	at	at	IN
hansen-can-2020	102	12	94.75	94.75	CD
hansen-can-2020	102	13	%	%	NN
hansen-can-2020	102	14	,	,	,
hansen-can-2020	102	15	the	the	DT
hansen-can-2020	102	16	article	article	NN
hansen-can-2020	102	17	categorization	categorization	NN
hansen-can-2020	102	18	at	at	IN
hansen-can-2020	102	19	81.2	81.2	CD
hansen-can-2020	102	20	%	%	NN
hansen-can-2020	102	21	,	,	,
hansen-can-2020	102	22	and	and	CC
hansen-can-2020	102	23	the	the	DT
hansen-can-2020	102	24	hierarchy	hierarchy	NN
hansen-can-2020	102	25	at	at	IN
hansen-can-2020	102	26	78	78	CD
hansen-can-2020	102	27	%	%	NN
hansen-can-2020	102	28	(	(	-LRB-
hansen-can-2020	102	29	Shen	Shen	NNP
hansen-can-2020	102	30	,	,	,
hansen-can-2020	102	31	Ma	Ma	NNP
hansen-can-2020	102	32	,	,	,
hansen-can-2020	102	33	and	and	CC
hansen-can-2020	102	34	Wang	Wang	NNP
hansen-can-2020	102	35	2018	2018	CD
hansen-can-2020	102	36	,	,	,
hansen-can-2020	102	37	5	5	CD
hansen-can-2020	102	38	)	)	-RRB-
hansen-can-2020	102	39	.	.	.
hansen-can-2020	103	1	Since	since	IN
hansen-can-2020	103	2	MSC	MSC	NNP
hansen-can-2020	103	3	was	be	VBD
hansen-can-2020	103	4	created	create	VBN
hansen-can-2020	103	5	by	by	IN
hansen-can-2020	103	6	humans	human	NNS
hansen-can-2020	103	7	there	there	EX
hansen-can-2020	103	8	is	be	VBZ
hansen-can-2020	103	9	no	no	DT
hansen-can-2020	103	10	meaningful	meaningful	JJ
hansen-can-2020	103	11	way	way	NN
hansen-can-2020	103	12	to	to	TO
hansen-can-2020	103	13	compare	compare	VB
hansen-can-2020	103	14	the	the	DT
hansen-can-2020	103	15	FoS	FoS	NNP
hansen-can-2020	103	16	accuracy	accuracy	NN
hansen-can-2020	103	17	measurements	measurement	NNS
hansen-can-2020	103	18	,	,	,
hansen-can-2020	103	19	but	but	CC
hansen-can-2020	103	20	the	the	DT
hansen-can-2020	103	21	categorization	categorization	NN
hansen-can-2020	103	22	accuracy	accuracy	NN
hansen-can-2020	103	23	falls	fall	VBZ
hansen-can-2020	103	24	somewhere	somewhere	RB
hansen-can-2020	103	25	between	between	IN
hansen-can-2020	103	26	that	that	DT
hansen-can-2020	103	27	of	of	IN
hansen-can-2020	103	28	the	the	DT
hansen-can-2020	103	29	two	two	CD
hansen-can-2020	103	30	mathematics	mathematic	NNS
hansen-can-2020	103	31	projects	project	NNS
hansen-can-2020	103	32	.	.	.
hansen-can-2020	104	1	This	this	DT
hansen-can-2020	104	2	is	be	VBZ
hansen-can-2020	104	3	a	a	DT
hansen-can-2020	104	4	very	very	RB
hansen-can-2020	104	5	impressive	impressive	JJ
hansen-can-2020	104	6	result	result	NN
hansen-can-2020	104	7	,	,	,
hansen-can-2020	104	8	especially	especially	RB
hansen-can-2020	104	9	when	when	WRB
hansen-can-2020	104	10	the	the	DT
hansen-can-2020	104	11	aforementioned	aforementioned	JJ
hansen-can-2020	104	12	scale	scale	NN
hansen-can-2020	104	13	is	be	VBZ
hansen-can-2020	104	14	taken	take	VBN
hansen-can-2020	104	15	into	into	IN
hansen-can-2020	104	16	account	account	NN
hansen-can-2020	104	17	.	.	.
hansen-can-2020	105	1	Instead	instead	RB
hansen-can-2020	105	2	of	of	IN
hansen-can-2020	105	3	trying	try	VBG
hansen-can-2020	105	4	to	to	TO
hansen-can-2020	105	5	replace	replace	VB
hansen-can-2020	105	6	the	the	DT
hansen-can-2020	105	7	work	work	NN
hansen-can-2020	105	8	of	of	IN
hansen-can-2020	105	9	humans	human	NNS
hansen-can-2020	105	10	categorizing	categorize	VBG
hansen-can-2020	105	11	mathematics	mathematic	NNS
hansen-can-2020	105	12	articles	article	NNS
hansen-can-2020	105	13	indexed	index	VBN
hansen-can-2020	105	14	in	in	IN
hansen-can-2020	105	15	a	a	DT
hansen-can-2020	105	16	database	database	NN
hansen-can-2020	105	17	,	,	,
hansen-can-2020	105	18	which	which	WDT
hansen-can-2020	105	19	for	for	IN
hansen-can-2020	105	20	2018	2018	CD
hansen-can-2020	105	21	was	be	VBD
hansen-can-2020	105	22	120,324	120,324	CD
hansen-can-2020	105	23	items	item	NNS
hansen-can-2020	105	24	in	in	IN
hansen-can-2020	105	25	MathSciNet[footnoteRef:17	mathscinet[footnoteref:17	CD
hansen-can-2020	105	26	]	]	-RRB-
hansen-can-2020	105	27	and	and	CC
hansen-can-2020	105	28	97,819	97,819	CD
hansen-can-2020	105	29	in	in	IN
hansen-can-2020	105	30	zbMath[footnoteRef:18	zbMath[footnoteRef:18	NNP
hansen-can-2020	105	31	]	]	-RRB-
hansen-can-2020	105	32	,	,	,
hansen-can-2020	105	33	it	-PRON-	PRP
hansen-can-2020	105	34	is	be	VBZ
hansen-can-2020	105	35	trying	try	VBG
hansen-can-2020	105	36	to	to	TO
hansen-can-2020	105	37	replace	replace	VB
hansen-can-2020	105	38	the	the	DT
hansen-can-2020	105	39	human	human	JJ
hansen-can-2020	105	40	categorization	categorization	NN
hansen-can-2020	105	41	of	of	IN
hansen-can-2020	105	42	all	all	DT
hansen-can-2020	105	43	items	item	NNS
hansen-can-2020	105	44	indexed	index	VBN
hansen-can-2020	105	45	in	in	IN
hansen-can-2020	105	46	MAG	MAG	NNP
hansen-can-2020	105	47	,	,	,
hansen-can-2020	105	48	which	which	WDT
hansen-can-2020	105	49	was	be	VBD
hansen-can-2020	105	50	10,616,601	10,616,601	CD
hansen-can-2020	105	51	in	in	IN
hansen-can-2020	105	52	2018	2018	CD
hansen-can-2020	105	53	.	.	.
hansen-can-2020	106	1	[	[	-LRB-
hansen-can-2020	106	2	footnoteRef:19	footnoteRef:19	NNP
hansen-can-2020	106	3	]	]	-RRB-
hansen-can-2020	106	4	Both	both	CC
hansen-can-2020	106	5	zbMath	zbMath	NNP
hansen-can-2020	106	6	and	and	CC
hansen-can-2020	106	7	MathSciNet	MathSciNet	NNP
hansen-can-2020	106	8	were	be	VBD
hansen-can-2020	106	9	capable	capable	JJ
hansen-can-2020	106	10	of	of	IN
hansen-can-2020	106	11	providing	provide	VBG
hansen-can-2020	106	12	the	the	DT
hansen-can-2020	106	13	human	human	JJ
hansen-can-2020	106	14	labor	labor	NN
hansen-can-2020	106	15	to	to	TO
hansen-can-2020	106	16	do	do	VB
hansen-can-2020	106	17	the	the	DT
hansen-can-2020	106	18	work	work	NN
hansen-can-2020	106	19	of	of	IN
hansen-can-2020	106	20	assigning	assign	VBG
hansen-can-2020	106	21	MSC	MSC	NNP
hansen-can-2020	106	22	values	value	NNS
hansen-can-2020	106	23	to	to	IN
hansen-can-2020	106	24	the	the	DT
hansen-can-2020	106	25	mathematics	mathematic	NNS
hansen-can-2020	106	26	articles	article	NNS
hansen-can-2020	106	27	they	-PRON-	PRP
hansen-can-2020	106	28	indexed	index	VBD
hansen-can-2020	106	29	in	in	IN
hansen-can-2020	106	30	2018	2018	CD
hansen-can-2020	106	31	.	.	.
hansen-can-2020	107	1	Therefore	therefore	RB
hansen-can-2020	107	2	using	use	VBG
hansen-can-2020	107	3	an	an	DT
hansen-can-2020	107	4	automated	automate	VBN
hansen-can-2020	107	5	categorization	categorization	NN
hansen-can-2020	107	6	,	,	,
hansen-can-2020	107	7	which	which	WDT
hansen-can-2020	107	8	at	at	IN
hansen-can-2020	107	9	best	good	JJS
hansen-can-2020	107	10	could	could	MD
hansen-can-2020	107	11	only	only	RB
hansen-can-2020	107	12	get	get	VB
hansen-can-2020	107	13	the	the	DT
hansen-can-2020	107	14	top	top	JJ
hansen-can-2020	107	15	level	level	NN
hansen-can-2020	107	16	right	right	RB
hansen-can-2020	107	17	with	with	IN
hansen-can-2020	107	18	90	90	CD
hansen-can-2020	107	19	%	%	NN
hansen-can-2020	107	20	accuracy	accuracy	NN
hansen-can-2020	107	21	,	,	,
hansen-can-2020	107	22	was	be	VBD
hansen-can-2020	107	23	not	not	RB
hansen-can-2020	107	24	the	the	DT
hansen-can-2020	107	25	right	right	JJ
hansen-can-2020	107	26	path	path	NN
hansen-can-2020	107	27	.	.	.
hansen-can-2020	108	1	On	on	IN
hansen-can-2020	108	2	the	the	DT
hansen-can-2020	108	3	other	other	JJ
hansen-can-2020	108	4	hand	hand	NN
hansen-can-2020	108	5	,	,	,
hansen-can-2020	108	6	it	-PRON-	PRP
hansen-can-2020	108	7	seems	seem	VBZ
hansen-can-2020	108	8	clear	clear	JJ
hansen-can-2020	108	9	that	that	IN
hansen-can-2020	108	10	no	no	DT
hansen-can-2020	108	11	one	one	PRP
hansen-can-2020	108	12	could	could	MD
hansen-can-2020	108	13	feasibly	feasibly	RB
hansen-can-2020	108	14	provide	provide	VB
hansen-can-2020	108	15	the	the	DT
hansen-can-2020	108	16	human	human	JJ
hansen-can-2020	108	17	labor	labor	NN
hansen-can-2020	108	18	to	to	TO
hansen-can-2020	108	19	categorize	categorize	VB
hansen-can-2020	108	20	all	all	DT
hansen-can-2020	108	21	articles	article	NNS
hansen-can-2020	108	22	indexed	index	VBN
hansen-can-2020	108	23	by	by	IN
hansen-can-2020	108	24	MAG	MAG	NNP
hansen-can-2020	108	25	in	in	IN
hansen-can-2020	108	26	2018	2018	CD
hansen-can-2020	108	27	so	so	IN
hansen-can-2020	108	28	an	an	DT
hansen-can-2020	108	29	80	80	CD
hansen-can-2020	108	30	%	%	NN
hansen-can-2020	108	31	accurate	accurate	JJ
hansen-can-2020	108	32	categorization	categorization	NN
hansen-can-2020	108	33	is	be	VBZ
hansen-can-2020	108	34	a	a	DT
hansen-can-2020	108	35	significant	significant	JJ
hansen-can-2020	108	36	accomplishment	accomplishment	NN
hansen-can-2020	108	37	.	.	.
hansen-can-2020	109	1	To	to	TO
hansen-can-2020	109	2	go	go	VB
hansen-can-2020	109	3	back	back	RB
hansen-can-2020	109	4	to	to	IN
hansen-can-2020	109	5	the	the	DT
hansen-can-2020	109	6	nail	nail	NN
hansen-can-2020	109	7	and	and	CC
hansen-can-2020	109	8	hammer	hammer	NN
hansen-can-2020	109	9	analogy	analogy	NN
hansen-can-2020	109	10	,	,	,
hansen-can-2020	109	11	Microsoft	Microsoft	NNP
hansen-can-2020	109	12	may	may	MD
hansen-can-2020	109	13	have	have	VB
hansen-can-2020	109	14	used	use	VBN
hansen-can-2020	109	15	a	a	DT
hansen-can-2020	109	16	sledgehammer	sledgehammer	NN
hansen-can-2020	109	17	but	but	CC
hansen-can-2020	109	18	they	-PRON-	PRP
hansen-can-2020	109	19	were	be	VBD
hansen-can-2020	109	20	hammering	hammer	VBG
hansen-can-2020	109	21	a	a	DT
hansen-can-2020	109	22	rather	rather	RB
hansen-can-2020	109	23	giant	giant	JJ
hansen-can-2020	109	24	nail	nail	NN
hansen-can-2020	109	25	.	.	.
hansen-can-2020	110	1	Comment	comment	NN
hansen-can-2020	110	2	by	by	IN
hansen-can-2020	110	3	Mark	Mark	NNP
hansen-can-2020	110	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	110	5	:	:	:
hansen-can-2020	110	6	Maybe	maybe	RB
hansen-can-2020	110	7	use	use	VB
hansen-can-2020	110	8	"	"	``
hansen-can-2020	110	9	the	the	DT
hansen-can-2020	110	10	project	project	NN
hansen-can-2020	110	11	?	?	.
hansen-can-2020	110	12	?	?	.
hansen-can-2020	111	1	Comment	comment	NN
hansen-can-2020	111	2	by	by	IN
hansen-can-2020	111	3	Mark	Mark	NNP
hansen-can-2020	111	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	111	5	:	:	:
hansen-can-2020	111	6	It	-PRON-	PRP
hansen-can-2020	111	7	would	would	MD
hansen-can-2020	111	8	be	be	VB
hansen-can-2020	111	9	interesting	interesting	JJ
hansen-can-2020	111	10	to	to	TO
hansen-can-2020	111	11	know	know	VB
hansen-can-2020	111	12	the	the	DT
hansen-can-2020	111	13	accuracy	accuracy	NN
hansen-can-2020	111	14	of	of	IN
hansen-can-2020	111	15	human	human	JJ
hansen-can-2020	111	16	indexing	indexing	NN
hansen-can-2020	111	17	in	in	IN
hansen-can-2020	111	18	MSN	MSN	NNP
hansen-can-2020	111	19	or	or	CC
hansen-can-2020	111	20	ZBM	ZBM	NNP
hansen-can-2020	111	21	.	.	.
hansen-can-2020	112	1	Comment	comment	NN
hansen-can-2020	112	2	by	by	IN
hansen-can-2020	112	3	Samuel	Samuel	NNP
hansen-can-2020	112	4	Hansen	Hansen	NNP
hansen-can-2020	112	5	:	:	:
hansen-can-2020	112	6	Sadly	sadly	RB
hansen-can-2020	112	7	no	no	DT
hansen-can-2020	112	8	one	one	NN
hansen-can-2020	112	9	seems	seem	VBZ
hansen-can-2020	112	10	to	to	TO
hansen-can-2020	112	11	have	have	VB
hansen-can-2020	112	12	studied	study	VBN
hansen-can-2020	112	13	this	this	DT
hansen-can-2020	112	14	,	,	,
hansen-can-2020	112	15	but	but	CC
hansen-can-2020	112	16	both	both	DT
hansen-can-2020	112	17	databases	database	NNS
hansen-can-2020	112	18	are	be	VBP
hansen-can-2020	112	19	considered	consider	VBN
hansen-can-2020	112	20	Gold	Gold	NNP
hansen-can-2020	112	21	Standard	Standard	NNP
hansen-can-2020	112	22	when	when	WRB
hansen-can-2020	112	23	it	-PRON-	PRP
hansen-can-2020	112	24	comes	come	VBZ
hansen-can-2020	112	25	to	to	IN
hansen-can-2020	112	26	classification	classification	NN
hansen-can-2020	112	27	.	.	.
hansen-can-2020	113	1	They	-PRON-	PRP
hansen-can-2020	113	2	do	do	VBP
hansen-can-2020	113	3	not	not	RB
hansen-can-2020	113	4	rely	rely	VB
hansen-can-2020	113	5	on	on	IN
hansen-can-2020	113	6	authors	author	NNS
hansen-can-2020	113	7	or	or	CC
hansen-can-2020	113	8	publishers	publisher	NNS
hansen-can-2020	113	9	providing	provide	VBG
hansen-can-2020	113	10	classification	classification	NN
hansen-can-2020	113	11	information	information	NN
hansen-can-2020	113	12	and	and	CC
hansen-can-2020	113	13	each	each	DT
hansen-can-2020	113	14	article	article	NN
hansen-can-2020	113	15	is	be	VBZ
hansen-can-2020	113	16	classified	classify	VBN
hansen-can-2020	113	17	not	not	RB
hansen-can-2020	113	18	only	only	RB
hansen-can-2020	113	19	by	by	IN
hansen-can-2020	113	20	the	the	DT
hansen-can-2020	113	21	subject	subject	JJ
hansen-can-2020	113	22	area	area	NN
hansen-can-2020	113	23	expert	expert	NN
hansen-can-2020	113	24	reviewer	reviewer	NN
hansen-can-2020	113	25	they	-PRON-	PRP
hansen-can-2020	113	26	send	send	VBP
hansen-can-2020	113	27	it	-PRON-	PRP
hansen-can-2020	113	28	to	to	IN
hansen-can-2020	113	29	it	-PRON-	PRP
hansen-can-2020	113	30	is	be	VBZ
hansen-can-2020	113	31	also	also	RB
hansen-can-2020	113	32	identified	identify	VBN
hansen-can-2020	113	33	before	before	IN
hansen-can-2020	113	34	the	the	DT
hansen-can-2020	113	35	reviewer	reviewer	NN
hansen-can-2020	113	36	receives	receive	VBZ
hansen-can-2020	113	37	it	-PRON-	PRP
hansen-can-2020	113	38	by	by	IN
hansen-can-2020	113	39	an	an	DT
hansen-can-2020	113	40	editor	editor	NN
hansen-can-2020	113	41	who	who	WP
hansen-can-2020	113	42	is	be	VBZ
hansen-can-2020	113	43	trying	try	VBG
hansen-can-2020	113	44	to	to	TO
hansen-can-2020	113	45	find	find	VB
hansen-can-2020	113	46	a	a	DT
hansen-can-2020	113	47	reviewer	reviewer	NN
hansen-can-2020	113	48	,	,	,
hansen-can-2020	113	49	and	and	CC
hansen-can-2020	113	50	again	again	RB
hansen-can-2020	113	51	after	after	IN
hansen-can-2020	113	52	the	the	DT
hansen-can-2020	113	53	paper	paper	NN
hansen-can-2020	113	54	is	be	VBZ
hansen-can-2020	113	55	reviewed	review	VBN
hansen-can-2020	113	56	as	as	IN
hansen-can-2020	113	57	a	a	DT
hansen-can-2020	113	58	check	check	NN
hansen-can-2020	113	59	.	.	.
hansen-can-2020	114	1	There	there	EX
hansen-can-2020	114	2	is	be	VBZ
hansen-can-2020	114	3	rarely	rarely	RB
hansen-can-2020	114	4	much	much	JJ
hansen-can-2020	114	5	disagreement	disagreement	NN
hansen-can-2020	114	6	over	over	IN
hansen-can-2020	114	7	MSC	MSC	NNP
hansen-can-2020	114	8	values	value	NNS
hansen-can-2020	114	9	between	between	IN
hansen-can-2020	114	10	MSN	MSN	NNP
hansen-can-2020	114	11	and	and	CC
hansen-can-2020	114	12	zbMath	zbMath	NNP
hansen-can-2020	114	13	and	and	CC
hansen-can-2020	114	14	when	when	WRB
hansen-can-2020	114	15	it	-PRON-	PRP
hansen-can-2020	114	16	does	do	VBZ
hansen-can-2020	114	17	occur	occur	VB
hansen-can-2020	114	18	,	,	,
hansen-can-2020	114	19	from	from	IN
hansen-can-2020	114	20	what	what	WP
hansen-can-2020	114	21	I	-PRON-	PRP
hansen-can-2020	114	22	am	be	VBP
hansen-can-2020	114	23	trying	try	VBG
hansen-can-2020	114	24	to	to	TO
hansen-can-2020	114	25	remember	remember	VB
hansen-can-2020	114	26	from	from	IN
hansen-can-2020	114	27	a	a	DT
hansen-can-2020	114	28	presentation	presentation	NN
hansen-can-2020	114	29	,	,	,
hansen-can-2020	114	30	it	-PRON-	PRP
hansen-can-2020	114	31	is	be	VBZ
hansen-can-2020	114	32	in	in	IN
hansen-can-2020	114	33	the	the	DT
hansen-can-2020	114	34	last	last	JJ
hansen-can-2020	114	35	part	part	NN
hansen-can-2020	114	36	so	so	IN
hansen-can-2020	114	37	we	-PRON-	PRP
hansen-can-2020	114	38	are	be	VBP
hansen-can-2020	114	39	taking	take	VBG
hansen-can-2020	114	40	just	just	RB
hansen-can-2020	114	41	tiny	tiny	JJ
hansen-can-2020	114	42	differences	difference	NNS
hansen-can-2020	114	43	given	give	VBN
hansen-can-2020	114	44	how	how	WRB
hansen-can-2020	114	45	fine	fine	JJ
hansen-can-2020	114	46	a	a	DT
hansen-can-2020	114	47	classification	classification	NN
hansen-can-2020	114	48	MSN	MSN	NNP
hansen-can-2020	114	49	is	be	VBZ
hansen-can-2020	114	50	.	.	.
hansen-can-2020	115	1	Comment	comment	NN
hansen-can-2020	115	2	by	by	IN
hansen-can-2020	115	3	Mark	Mark	NNP
hansen-can-2020	115	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	115	5	:	:	:
hansen-can-2020	115	6	If	if	IN
hansen-can-2020	115	7	you	-PRON-	PRP
hansen-can-2020	115	8	can	can	MD
hansen-can-2020	115	9	find	find	VB
hansen-can-2020	115	10	it	-PRON-	PRP
hansen-can-2020	115	11	,	,	,
hansen-can-2020	115	12	it	-PRON-	PRP
hansen-can-2020	115	13	may	may	MD
hansen-can-2020	115	14	be	be	VB
hansen-can-2020	115	15	worth	worth	JJ
hansen-can-2020	115	16	citing	cite	VBG
hansen-can-2020	115	17	OR	or	CC
hansen-can-2020	115	18	you	-PRON-	PRP
hansen-can-2020	115	19	could	could	MD
hansen-can-2020	115	20	explain	explain	VB
hansen-can-2020	115	21	their	-PRON-	PRP$
hansen-can-2020	115	22	methodology	methodology	NN
hansen-can-2020	115	23	.	.	.
hansen-can-2020	116	1	It	-PRON-	PRP
hansen-can-2020	116	2	helps	help	VBZ
hansen-can-2020	116	3	validate	validate	VB
hansen-can-2020	116	4	the	the	DT
hansen-can-2020	116	5	claim	claim	NN
hansen-can-2020	116	6	that	that	IN
hansen-can-2020	116	7	human	human	JJ
hansen-can-2020	116	8	categorization	categorization	NN
hansen-can-2020	116	9	is	be	VBZ
hansen-can-2020	116	10	better	well	JJR
hansen-can-2020	116	11	than	than	IN
hansen-can-2020	116	12	machine	machine	NN
hansen-can-2020	116	13	.	.	.
hansen-can-2020	117	1	Comment	comment	NN
hansen-can-2020	117	2	by	by	IN
hansen-can-2020	117	3	Daniel	Daniel	NNP
hansen-can-2020	117	4	Johnson	Johnson	NNP
hansen-can-2020	117	5	:	:	:
hansen-can-2020	117	6	This	this	DT
hansen-can-2020	117	7	is	be	VBZ
hansen-can-2020	117	8	kind	kind	RB
hansen-can-2020	117	9	of	of	RB
hansen-can-2020	117	10	interesting	interesting	JJ
hansen-can-2020	117	11	,	,	,
hansen-can-2020	117	12	as	as	IN
hansen-can-2020	117	13	it	-PRON-	PRP
hansen-can-2020	117	14	might	may	MD
hansen-can-2020	117	15	be	be	VB
hansen-can-2020	117	16	a	a	DT
hansen-can-2020	117	17	method	method	NN
hansen-can-2020	117	18	of	of	IN
hansen-can-2020	117	19	working	work	VBG
hansen-can-2020	117	20	other	other	JJ
hansen-can-2020	117	21	disciplines	discipline	NNS
hansen-can-2020	117	22	have	have	VBP
hansen-can-2020	117	23	never	never	RB
hansen-can-2020	117	24	heard	hear	VBN
hansen-can-2020	117	25	of	of	IN
hansen-can-2020	117	26	.	.	.
hansen-can-2020	118	1	Perhaps	perhaps	RB
hansen-can-2020	118	2	a	a	DT
hansen-can-2020	118	3	short	short	JJ
hansen-can-2020	118	4	footnote	footnote	NN
hansen-can-2020	118	5	?	?	.
hansen-can-2020	119	1	Comment	comment	NN
hansen-can-2020	119	2	by	by	IN
hansen-can-2020	119	3	Mark	Mark	NNP
hansen-can-2020	119	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	119	5	:	:	:
hansen-can-2020	119	6	Maybe	maybe	RB
hansen-can-2020	119	7	"	"	``
hansen-can-2020	119	8	approach	approach	NN
hansen-can-2020	119	9	"	"	''
hansen-can-2020	119	10	?	?	.
hansen-can-2020	120	1	[	[	-LRB-
hansen-can-2020	120	2	17	17	CD
hansen-can-2020	120	3	:	:	:
hansen-can-2020	120	4	https://mathscinet.ams.org/mathscinet/search/publications.html?dr=pubyear&yrop=eq&arg3=2018	https://mathscinet.ams.org/mathscinet/search/publications.html?dr=pubyear&yrop=eq&arg3=2018	NNP
hansen-can-2020	120	5	]	]	-RRB-
hansen-can-2020	120	6	[	[	-LRB-
hansen-can-2020	120	7	18	18	CD
hansen-can-2020	120	8	:	:	:
hansen-can-2020	120	9	https://zbmath.org/?q=py%3A2018	https://zbmath.org/?q=py%3a2018	LS
hansen-can-2020	120	10	]	]	-RRB-
hansen-can-2020	120	11	[	[	-LRB-
hansen-can-2020	120	12	19	19	CD
hansen-can-2020	120	13	:	:	:
hansen-can-2020	120	14	https://academic.microsoft.com/publications/33923547	https://academic.microsoft.com/publications/33923547	LS
hansen-can-2020	120	15	]	]	-RRB-
hansen-can-2020	120	16	Are	be	VBP
hansen-can-2020	120	17	You	-PRON-	PRP
hansen-can-2020	120	18	Sure	sure	JJ
hansen-can-2020	120	19	it	-PRON-	PRP
hansen-can-2020	120	20	’s	’	VBZ
hansen-can-2020	120	21	a	a	DT
hansen-can-2020	120	22	Nail	nail	NN
hansen-can-2020	120	23	?	?	.
hansen-can-2020	121	1	I	-PRON-	PRP
hansen-can-2020	121	2	started	start	VBD
hansen-can-2020	121	3	this	this	DT
hansen-can-2020	121	4	chapter	chapter	NN
hansen-can-2020	121	5	talking	talk	VBG
hansen-can-2020	121	6	about	about	IN
hansen-can-2020	121	7	how	how	WRB
hansen-can-2020	121	8	we	-PRON-	PRP
hansen-can-2020	121	9	have	have	VBP
hansen-can-2020	121	10	all	all	DT
hansen-can-2020	121	11	been	be	VBN
hansen-can-2020	121	12	told	tell	VBN
hansen-can-2020	121	13	that	that	IN
hansen-can-2020	121	14	AI	AI	NNP
hansen-can-2020	121	15	and	and	CC
hansen-can-2020	121	16	machine	machine	NN
hansen-can-2020	121	17	learning	learning	NN
hansen-can-2020	121	18	were	be	VBD
hansen-can-2020	121	19	going	go	VBG
hansen-can-2020	121	20	to	to	TO
hansen-can-2020	121	21	revolutionize	revolutionize	VB
hansen-can-2020	121	22	everything	everything	NN
hansen-can-2020	121	23	in	in	IN
hansen-can-2020	121	24	the	the	DT
hansen-can-2020	121	25	world	world	NN
hansen-can-2020	121	26	.	.	.
hansen-can-2020	122	1	That	that	IN
hansen-can-2020	122	2	they	-PRON-	PRP
hansen-can-2020	122	3	were	be	VBD
hansen-can-2020	122	4	the	the	DT
hansen-can-2020	122	5	hammers	hammer	NNS
hansen-can-2020	122	6	and	and	CC
hansen-can-2020	122	7	all	all	PDT
hansen-can-2020	122	8	the	the	DT
hansen-can-2020	122	9	world	world	NN
hansen-can-2020	122	10	's	's	POS
hansen-can-2020	122	11	problems	problem	NNS
hansen-can-2020	122	12	were	be	VBD
hansen-can-2020	122	13	nails	nail	NNS
hansen-can-2020	122	14	.	.	.
hansen-can-2020	123	1	I	-PRON-	PRP
hansen-can-2020	123	2	found	find	VBD
hansen-can-2020	123	3	that	that	IN
hansen-can-2020	123	4	this	this	DT
hansen-can-2020	123	5	was	be	VBD
hansen-can-2020	123	6	not	not	RB
hansen-can-2020	123	7	the	the	DT
hansen-can-2020	123	8	case	case	NN
hansen-can-2020	123	9	when	when	WRB
hansen-can-2020	123	10	we	-PRON-	PRP
hansen-can-2020	123	11	tried	try	VBD
hansen-can-2020	123	12	to	to	TO
hansen-can-2020	123	13	employ	employ	VB
hansen-can-2020	123	14	it	-PRON-	PRP
hansen-can-2020	123	15	,	,	,
hansen-can-2020	123	16	in	in	IN
hansen-can-2020	123	17	an	an	DT
hansen-can-2020	123	18	admittedly	admittedly	RB
hansen-can-2020	123	19	rather	rather	RB
hansen-can-2020	123	20	naive	naive	JJ
hansen-can-2020	123	21	fashion	fashion	NN
hansen-can-2020	123	22	,	,	,
hansen-can-2020	123	23	to	to	TO
hansen-can-2020	123	24	automatically	automatically	RB
hansen-can-2020	123	25	categorize	categorize	VB
hansen-can-2020	123	26	mathematical	mathematical	JJ
hansen-can-2020	123	27	articles	article	NNS
hansen-can-2020	123	28	.	.	.
hansen-can-2020	124	1	From	from	IN
hansen-can-2020	124	2	the	the	DT
hansen-can-2020	124	3	other	other	JJ
hansen-can-2020	124	4	examples	example	NNS
hansen-can-2020	124	5	I	-PRON-	PRP
hansen-can-2020	124	6	included	include	VBD
hansen-can-2020	124	7	,	,	,
hansen-can-2020	124	8	it	-PRON-	PRP
hansen-can-2020	124	9	is	be	VBZ
hansen-can-2020	124	10	also	also	RB
hansen-can-2020	124	11	clear	clear	JJ
hansen-can-2020	124	12	computational	computational	JJ
hansen-can-2020	124	13	experts	expert	NNS
hansen-can-2020	124	14	find	find	VBP
hansen-can-2020	124	15	the	the	DT
hansen-can-2020	124	16	automatic	automatic	JJ
hansen-can-2020	124	17	categorization	categorization	NN
hansen-can-2020	124	18	of	of	IN
hansen-can-2020	124	19	highly	highly	RB
hansen-can-2020	124	20	technical	technical	JJ
hansen-can-2020	124	21	content	content	NN
hansen-can-2020	124	22	a	a	DT
hansen-can-2020	124	23	hard	hard	JJ
hansen-can-2020	124	24	problem	problem	NN
hansen-can-2020	124	25	to	to	TO
hansen-can-2020	124	26	tackle	tackle	VB
hansen-can-2020	124	27	,	,	,
hansen-can-2020	124	28	one	one	CD
hansen-can-2020	124	29	where	where	WRB
hansen-can-2020	124	30	success	success	NN
hansen-can-2020	124	31	is	be	VBZ
hansen-can-2020	124	32	very	very	RB
hansen-can-2020	124	33	much	much	RB
hansen-can-2020	124	34	dependent	dependent	JJ
hansen-can-2020	124	35	on	on	IN
hansen-can-2020	124	36	what	what	WP
hansen-can-2020	124	37	it	-PRON-	PRP
hansen-can-2020	124	38	is	be	VBZ
hansen-can-2020	124	39	being	be	VBG
hansen-can-2020	124	40	measured	measure	VBN
hansen-can-2020	124	41	against	against	IN
hansen-can-2020	124	42	.	.	.
hansen-can-2020	125	1	In	in	IN
hansen-can-2020	125	2	the	the	DT
hansen-can-2020	125	3	case	case	NN
hansen-can-2020	125	4	of	of	IN
hansen-can-2020	125	5	classifying	classify	VBG
hansen-can-2020	125	6	mathematics	mathematic	NNS
hansen-can-2020	125	7	,	,	,
hansen-can-2020	125	8	machine	machine	NN
hansen-can-2020	125	9	learning	learning	NN
hansen-can-2020	125	10	can	can	MD
hansen-can-2020	125	11	do	do	VB
hansen-can-2020	125	12	a	a	DT
hansen-can-2020	125	13	decent	decent	JJ
hansen-can-2020	125	14	job	job	NN
hansen-can-2020	125	15	but	but	CC
hansen-can-2020	125	16	not	not	RB
hansen-can-2020	125	17	enough	enough	JJ
hansen-can-2020	125	18	to	to	TO
hansen-can-2020	125	19	compete	compete	VB
hansen-can-2020	125	20	with	with	IN
hansen-can-2020	125	21	humans	human	NNS
hansen-can-2020	125	22	.	.	.
hansen-can-2020	126	1	In	in	IN
hansen-can-2020	126	2	the	the	DT
hansen-can-2020	126	3	case	case	NN
hansen-can-2020	126	4	of	of	IN
hansen-can-2020	126	5	classifying	classify	VBG
hansen-can-2020	126	6	everything	everything	NN
hansen-can-2020	126	7	,	,	,
hansen-can-2020	126	8	scale	scale	NN
hansen-can-2020	126	9	gives	give	VBZ
hansen-can-2020	126	10	machines	machine	NNS
hansen-can-2020	126	11	an	an	DT
hansen-can-2020	126	12	edge	edge	NN
hansen-can-2020	126	13	,	,	,
hansen-can-2020	126	14	as	as	RB
hansen-can-2020	126	15	long	long	RB
hansen-can-2020	126	16	as	as	IN
hansen-can-2020	126	17	you	-PRON-	PRP
hansen-can-2020	126	18	have	have	VBP
hansen-can-2020	126	19	the	the	DT
hansen-can-2020	126	20	computational	computational	JJ
hansen-can-2020	126	21	power	power	NN
hansen-can-2020	126	22	and	and	CC
hansen-can-2020	126	23	knowledge	knowledge	NN
hansen-can-2020	126	24	wielded	wield	VBN
hansen-can-2020	126	25	by	by	IN
hansen-can-2020	126	26	a	a	DT
hansen-can-2020	126	27	company	company	NN
hansen-can-2020	126	28	like	like	IN
hansen-can-2020	126	29	Microsoft	Microsoft	NNP
hansen-can-2020	126	30	.	.	.
hansen-can-2020	127	1	Comment	comment	NN
hansen-can-2020	127	2	by	by	IN
hansen-can-2020	127	3	Mark	Mark	NNP
hansen-can-2020	127	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	127	5	:	:	:
hansen-can-2020	127	6	This	this	DT
hansen-can-2020	127	7	is	be	VBZ
hansen-can-2020	127	8	a	a	DT
hansen-can-2020	127	9	bit	bit	NN
hansen-can-2020	127	10	hyperbolic	hyperbolic	JJ
hansen-can-2020	127	11	-	-	:
hansen-can-2020	127	12	which	which	WDT
hansen-can-2020	127	13	is	be	VBZ
hansen-can-2020	127	14	okay	okay	JJ
hansen-can-2020	127	15	if	if	IN
hansen-can-2020	127	16	the	the	DT
hansen-can-2020	127	17	tone	tone	NN
hansen-can-2020	127	18	is	be	VBZ
hansen-can-2020	127	19	intended	intend	VBN
hansen-can-2020	127	20	.	.	.
hansen-can-2020	128	1	Certainly	certainly	RB
hansen-can-2020	128	2	,	,	,
hansen-can-2020	128	3	it	-PRON-	PRP
hansen-can-2020	128	4	would	would	MD
hansen-can-2020	128	5	be	be	VB
hansen-can-2020	128	6	easy	easy	JJ
hansen-can-2020	128	7	to	to	TO
hansen-can-2020	128	8	claim	claim	VB
hansen-can-2020	128	9	that	that	IN
hansen-can-2020	128	10	ML	ML	NNP
hansen-can-2020	128	11	intended	intend	VBD
hansen-can-2020	128	12	to	to	TO
hansen-can-2020	128	13	revolutionize	revolutionize	VB
hansen-can-2020	128	14	classification	classification	NN
hansen-can-2020	128	15	.	.	.
hansen-can-2020	129	1	This	this	DT
hansen-can-2020	129	2	collection	collection	NN
hansen-can-2020	129	3	is	be	VBZ
hansen-can-2020	129	4	about	about	IN
hansen-can-2020	129	5	the	the	DT
hansen-can-2020	129	6	intersection	intersection	NN
hansen-can-2020	129	7	of	of	IN
hansen-can-2020	129	8	AI	AI	NNP
hansen-can-2020	129	9	,	,	,
hansen-can-2020	129	10	machine	machine	NN
hansen-can-2020	129	11	learning	learning	NN
hansen-can-2020	129	12	,	,	,
hansen-can-2020	129	13	deep	deep	JJ
hansen-can-2020	129	14	learning	learning	NN
hansen-can-2020	129	15	,	,	,
hansen-can-2020	129	16	and	and	CC
hansen-can-2020	129	17	libraries	library	NNS
hansen-can-2020	129	18	.	.	.
hansen-can-2020	130	1	While	while	IN
hansen-can-2020	130	2	there	there	EX
hansen-can-2020	130	3	are	be	VBP
hansen-can-2020	130	4	definitely	definitely	RB
hansen-can-2020	130	5	problems	problem	NNS
hansen-can-2020	130	6	in	in	IN
hansen-can-2020	130	7	libraries	library	NNS
hansen-can-2020	130	8	where	where	WRB
hansen-can-2020	130	9	these	these	DT
hansen-can-2020	130	10	techniques	technique	NNS
hansen-can-2020	130	11	will	will	MD
hansen-can-2020	130	12	be	be	VB
hansen-can-2020	130	13	the	the	DT
hansen-can-2020	130	14	answer	answer	NN
hansen-can-2020	130	15	,	,	,
hansen-can-2020	130	16	I	-PRON-	PRP
hansen-can-2020	130	17	think	think	VBP
hansen-can-2020	130	18	it	-PRON-	PRP
hansen-can-2020	130	19	is	be	VBZ
hansen-can-2020	130	20	important	important	JJ
hansen-can-2020	130	21	to	to	TO
hansen-can-2020	130	22	pause	pause	VB
hansen-can-2020	130	23	,	,	,
hansen-can-2020	130	24	think	think	VB
hansen-can-2020	130	25	,	,	,
hansen-can-2020	130	26	and	and	CC
hansen-can-2020	130	27	be	be	VB
hansen-can-2020	130	28	sure	sure	JJ
hansen-can-2020	130	29	artificial	artificial	JJ
hansen-can-2020	130	30	intelligence	intelligence	NN
hansen-can-2020	130	31	techniques	technique	NNS
hansen-can-2020	130	32	are	be	VBP
hansen-can-2020	130	33	the	the	DT
hansen-can-2020	130	34	most	most	RBS
hansen-can-2020	130	35	effective	effective	JJ
hansen-can-2020	130	36	approach	approach	NN
hansen-can-2020	130	37	before	before	IN
hansen-can-2020	130	38	trying	try	VBG
hansen-can-2020	130	39	to	to	TO
hansen-can-2020	130	40	use	use	VB
hansen-can-2020	130	41	them	-PRON-	PRP
hansen-can-2020	130	42	.	.	.
hansen-can-2020	131	1	Libraries	library	NNS
hansen-can-2020	131	2	,	,	,
hansen-can-2020	131	3	even	even	RB
hansen-can-2020	131	4	those	those	DT
hansen-can-2020	131	5	like	like	IN
hansen-can-2020	131	6	the	the	DT
hansen-can-2020	131	7	one	one	NN
hansen-can-2020	131	8	I	-PRON-	PRP
hansen-can-2020	131	9	work	work	VBP
hansen-can-2020	131	10	in	in	RP
hansen-can-2020	131	11	,	,	,
hansen-can-2020	131	12	which	which	WDT
hansen-can-2020	131	13	are	be	VBP
hansen-can-2020	131	14	lucky	lucky	JJ
hansen-can-2020	131	15	enough	enough	RB
hansen-can-2020	131	16	to	to	TO
hansen-can-2020	131	17	boast	boast	VB
hansen-can-2020	131	18	of	of	IN
hansen-can-2020	131	19	incredibly	incredibly	RB
hansen-can-2020	131	20	talented	talented	JJ
hansen-can-2020	131	21	IT	IT	NNP
hansen-can-2020	131	22	departments	department	NNS
hansen-can-2020	131	23	,	,	,
hansen-can-2020	131	24	do	do	VB
hansen-can-2020	131	25	not	not	RB
hansen-can-2020	131	26	tend	tend	VB
hansen-can-2020	131	27	to	to	TO
hansen-can-2020	131	28	have	have	VB
hansen-can-2020	131	29	access	access	NN
hansen-can-2020	131	30	to	to	IN
hansen-can-2020	131	31	a	a	DT
hansen-can-2020	131	32	large	large	JJ
hansen-can-2020	131	33	amount	amount	NN
hansen-can-2020	131	34	of	of	IN
hansen-can-2020	131	35	unused	unused	JJ
hansen-can-2020	131	36	computational	computational	JJ
hansen-can-2020	131	37	power	power	NN
hansen-can-2020	131	38	or	or	CC
hansen-can-2020	131	39	numerous	numerous	JJ
hansen-can-2020	131	40	experts	expert	NNS
hansen-can-2020	131	41	in	in	IN
hansen-can-2020	131	42	bleeding	bleed	VBG
hansen-can-2020	131	43	-	-	HYPH
hansen-can-2020	131	44	edge	edge	NN
hansen-can-2020	131	45	AI	ai	NN
hansen-can-2020	131	46	.	.	.
hansen-can-2020	132	1	They	-PRON-	PRP
hansen-can-2020	132	2	are	be	VBP
hansen-can-2020	132	3	also	also	RB
hansen-can-2020	132	4	rather	rather	RB
hansen-can-2020	132	5	notoriously	notoriously	RB
hansen-can-2020	132	6	limited	limited	JJ
hansen-can-2020	132	7	budget	budget	NN
hansen-can-2020	132	8	-	-	HYPH
hansen-can-2020	132	9	wise	wise	JJ
hansen-can-2020	132	10	and	and	CC
hansen-can-2020	132	11	would	would	MD
hansen-can-2020	132	12	likely	likely	RB
hansen-can-2020	132	13	have	have	VB
hansen-can-2020	132	14	to	to	TO
hansen-can-2020	132	15	decide	decide	VB
hansen-can-2020	132	16	between	between	IN
hansen-can-2020	132	17	existing	exist	VBG
hansen-can-2020	132	18	budget	budget	NN
hansen-can-2020	132	19	items	item	NNS
hansen-can-2020	132	20	and	and	CC
hansen-can-2020	132	21	developing	develop	VBG
hansen-can-2020	132	22	an	an	DT
hansen-can-2020	132	23	in	in	IN
hansen-can-2020	132	24	-	-	HYPH
hansen-can-2020	132	25	house	house	NN
hansen-can-2020	132	26	machine	machine	NN
hansen-can-2020	132	27	learning	learning	NN
hansen-can-2020	132	28	program	program	NN
hansen-can-2020	132	29	.	.	.
hansen-can-2020	133	1	Those	those	DT
hansen-can-2020	133	2	realities	reality	NNS
hansen-can-2020	133	3	combined	combine	VBN
hansen-can-2020	133	4	with	with	IN
hansen-can-2020	133	5	the	the	DT
hansen-can-2020	133	6	legitimate	legitimate	JJ
hansen-can-2020	133	7	questions	question	NNS
hansen-can-2020	133	8	which	which	WDT
hansen-can-2020	133	9	can	can	MD
hansen-can-2020	133	10	be	be	VB
hansen-can-2020	133	11	raised	raise	VBN
hansen-can-2020	133	12	about	about	IN
hansen-can-2020	133	13	the	the	DT
hansen-can-2020	133	14	efficacy	efficacy	NN
hansen-can-2020	133	15	of	of	IN
hansen-can-2020	133	16	machine	machine	NN
hansen-can-2020	133	17	learning	learning	NN
hansen-can-2020	133	18	and	and	CC
hansen-can-2020	133	19	AI	AI	NNP
hansen-can-2020	133	20	with	with	IN
hansen-can-2020	133	21	respect	respect	NN
hansen-can-2020	133	22	to	to	IN
hansen-can-2020	133	23	the	the	DT
hansen-can-2020	133	24	types	type	NNS
hansen-can-2020	133	25	of	of	IN
hansen-can-2020	133	26	problems	problem	NNS
hansen-can-2020	133	27	a	a	DT
hansen-can-2020	133	28	library	library	NN
hansen-can-2020	133	29	may	may	MD
hansen-can-2020	133	30	encounter	encounter	VB
hansen-can-2020	133	31	,	,	,
hansen-can-2020	133	32	such	such	JJ
hansen-can-2020	133	33	as	as	IN
hansen-can-2020	133	34	categorizing	categorize	VBG
hansen-can-2020	133	35	the	the	DT
hansen-can-2020	133	36	contents	content	NNS
hansen-can-2020	133	37	of	of	IN
hansen-can-2020	133	38	highly	highly	RB
hansen-can-2020	133	39	technical	technical	JJ
hansen-can-2020	133	40	articles	article	NNS
hansen-can-2020	133	41	,	,	,
hansen-can-2020	133	42	make	make	VB
hansen-can-2020	133	43	me	-PRON-	PRP
hansen-can-2020	133	44	worry	worry	VB
hansen-can-2020	133	45	.	.	.
hansen-can-2020	134	1	While	while	IN
hansen-can-2020	134	2	there	there	EX
hansen-can-2020	134	3	will	will	MD
hansen-can-2020	134	4	be	be	VB
hansen-can-2020	134	5	many	many	JJ
hansen-can-2020	134	6	cases	case	NNS
hansen-can-2020	134	7	where	where	WRB
hansen-can-2020	134	8	using	use	VBG
hansen-can-2020	134	9	AI	AI	NNP
hansen-can-2020	134	10	makes	make	VBZ
hansen-can-2020	134	11	sense	sense	NN
hansen-can-2020	134	12	,	,	,
hansen-can-2020	134	13	I	-PRON-	PRP
hansen-can-2020	134	14	want	want	VBP
hansen-can-2020	134	15	to	to	TO
hansen-can-2020	134	16	be	be	VB
hansen-can-2020	134	17	sure	sure	JJ
hansen-can-2020	134	18	libraries	library	NNS
hansen-can-2020	134	19	are	be	VBP
hansen-can-2020	134	20	asking	ask	VBG
hansen-can-2020	134	21	themselves	-PRON-	PRP
hansen-can-2020	134	22	a	a	DT
hansen-can-2020	134	23	lot	lot	NN
hansen-can-2020	134	24	of	of	IN
hansen-can-2020	134	25	questions	question	NNS
hansen-can-2020	134	26	before	before	IN
hansen-can-2020	134	27	starting	start	VBG
hansen-can-2020	134	28	to	to	TO
hansen-can-2020	134	29	use	use	VB
hansen-can-2020	134	30	it	-PRON-	PRP
hansen-can-2020	134	31	.	.	.
hansen-can-2020	135	1	Questions	question	NNS
hansen-can-2020	135	2	like	like	IN
hansen-can-2020	135	3	:	:	:
hansen-can-2020	135	4	is	be	VBZ
hansen-can-2020	135	5	this	this	DT
hansen-can-2020	135	6	problem	problem	NN
hansen-can-2020	135	7	large	large	JJ
hansen-can-2020	135	8	enough	enough	RB
hansen-can-2020	135	9	in	in	IN
hansen-can-2020	135	10	scale	scale	NN
hansen-can-2020	135	11	to	to	IN
hansen-can-2020	135	12	substitute	substitute	NN
hansen-can-2020	135	13	machines	machine	NNS
hansen-can-2020	135	14	for	for	IN
hansen-can-2020	135	15	human	human	JJ
hansen-can-2020	135	16	labor	labor	NN
hansen-can-2020	135	17	given	give	VBN
hansen-can-2020	135	18	that	that	IN
hansen-can-2020	135	19	machines	machine	NNS
hansen-can-2020	135	20	will	will	MD
hansen-can-2020	135	21	likely	likely	RB
hansen-can-2020	135	22	be	be	VB
hansen-can-2020	135	23	less	less	RBR
hansen-can-2020	135	24	accurate	accurate	JJ
hansen-can-2020	135	25	?	?	.
hansen-can-2020	136	1	Or	or	CC
hansen-can-2020	136	2	:	:	:
hansen-can-2020	136	3	will	will	MD
hansen-can-2020	136	4	using	use	VBG
hansen-can-2020	136	5	machines	machine	NNS
hansen-can-2020	136	6	to	to	TO
hansen-can-2020	136	7	solve	solve	VB
hansen-can-2020	136	8	this	this	DT
hansen-can-2020	136	9	problem	problem	NN
hansen-can-2020	136	10	cost	cost	VB
hansen-can-2020	136	11	us	-PRON-	PRP
hansen-can-2020	136	12	more	more	RBR
hansen-can-2020	136	13	in	in	IN
hansen-can-2020	136	14	equipment	equipment	NN
hansen-can-2020	136	15	and	and	CC
hansen-can-2020	136	16	highly	highly	RB
hansen-can-2020	136	17	technical	technical	JJ
hansen-can-2020	136	18	staff	staff	NN
hansen-can-2020	136	19	than	than	IN
hansen-can-2020	136	20	our	-PRON-	PRP$
hansen-can-2020	136	21	current	current	JJ
hansen-can-2020	136	22	solution	solution	NN
hansen-can-2020	136	23	,	,	,
hansen-can-2020	136	24	and	and	CC
hansen-can-2020	136	25	has	have	VBZ
hansen-can-2020	136	26	that	that	DT
hansen-can-2020	136	27	factored	factor	VBN
hansen-can-2020	136	28	in	in	IN
hansen-can-2020	136	29	the	the	DT
hansen-can-2020	136	30	people	people	NNS
hansen-can-2020	136	31	and	and	CC
hansen-can-2020	136	32	services	service	NNS
hansen-can-2020	136	33	a	a	DT
hansen-can-2020	136	34	library	library	NN
hansen-can-2020	136	35	may	may	MD
hansen-can-2020	136	36	need	need	VB
hansen-can-2020	136	37	to	to	TO
hansen-can-2020	136	38	cut	cut	VB
hansen-can-2020	136	39	to	to	TO
hansen-can-2020	136	40	afford	afford	VB
hansen-can-2020	136	41	them	-PRON-	PRP
hansen-can-2020	136	42	?	?	.
hansen-can-2020	137	1	Not	not	RB
hansen-can-2020	137	2	to	to	TO
hansen-can-2020	137	3	mention	mention	VB
hansen-can-2020	137	4	:	:	:
hansen-can-2020	137	5	is	be	VBZ
hansen-can-2020	137	6	this	this	DT
hansen-can-2020	137	7	really	really	RB
hansen-can-2020	137	8	a	a	DT
hansen-can-2020	137	9	problem	problem	NN
hansen-can-2020	137	10	or	or	CC
hansen-can-2020	137	11	are	be	VBP
hansen-can-2020	137	12	we	-PRON-	PRP
hansen-can-2020	137	13	just	just	RB
hansen-can-2020	137	14	looking	look	VBG
hansen-can-2020	137	15	for	for	IN
hansen-can-2020	137	16	a	a	DT
hansen-can-2020	137	17	way	way	NN
hansen-can-2020	137	18	to	to	TO
hansen-can-2020	137	19	employ	employ	VB
hansen-can-2020	137	20	machine	machine	NN
hansen-can-2020	137	21	learning	learn	VBG
hansen-can-2020	137	22	to	to	TO
hansen-can-2020	137	23	say	say	VB
hansen-can-2020	137	24	that	that	IN
hansen-can-2020	137	25	we	-PRON-	PRP
hansen-can-2020	137	26	did	do	VBD
hansen-can-2020	137	27	?	?	.
hansen-can-2020	138	1	In	in	IN
hansen-can-2020	138	2	the	the	DT
hansen-can-2020	138	3	cases	case	NNS
hansen-can-2020	138	4	where	where	WRB
hansen-can-2020	138	5	the	the	DT
hansen-can-2020	138	6	answers	answer	NNS
hansen-can-2020	138	7	to	to	IN
hansen-can-2020	138	8	these	these	DT
hansen-can-2020	138	9	questions	question	NNS
hansen-can-2020	138	10	are	be	VBP
hansen-can-2020	138	11	yes	yes	UH
hansen-can-2020	138	12	,	,	,
hansen-can-2020	138	13	it	-PRON-	PRP
hansen-can-2020	138	14	will	will	MD
hansen-can-2020	138	15	make	make	VB
hansen-can-2020	138	16	sense	sense	NN
hansen-can-2020	138	17	for	for	IN
hansen-can-2020	138	18	libraries	library	NNS
hansen-can-2020	138	19	to	to	TO
hansen-can-2020	138	20	employ	employ	VB
hansen-can-2020	138	21	machine	machine	NN
hansen-can-2020	138	22	learning	learning	NN
hansen-can-2020	138	23	.	.	.
hansen-can-2020	139	1	I	-PRON-	PRP
hansen-can-2020	139	2	just	just	RB
hansen-can-2020	139	3	want	want	VBP
hansen-can-2020	139	4	libraries	library	NNS
hansen-can-2020	139	5	to	to	TO
hansen-can-2020	139	6	look	look	VB
hansen-can-2020	139	7	really	really	RB
hansen-can-2020	139	8	carefully	carefully	RB
hansen-can-2020	139	9	at	at	IN
hansen-can-2020	139	10	how	how	WRB
hansen-can-2020	139	11	they	-PRON-	PRP
hansen-can-2020	139	12	approach	approach	VBP
hansen-can-2020	139	13	problems	problem	NNS
hansen-can-2020	139	14	and	and	CC
hansen-can-2020	139	15	solutions	solution	NNS
hansen-can-2020	139	16	,	,	,
hansen-can-2020	139	17	to	to	TO
hansen-can-2020	139	18	make	make	VB
hansen-can-2020	139	19	sure	sure	JJ
hansen-can-2020	139	20	that	that	IN
hansen-can-2020	139	21	their	-PRON-	PRP$
hansen-can-2020	139	22	problem	problem	NN
hansen-can-2020	139	23	is	be	VBZ
hansen-can-2020	139	24	,	,	,
hansen-can-2020	139	25	in	in	IN
hansen-can-2020	139	26	fact	fact	NN
hansen-can-2020	139	27	,	,	,
hansen-can-2020	139	28	a	a	DT
hansen-can-2020	139	29	nail	nail	NN
hansen-can-2020	139	30	,	,	,
hansen-can-2020	139	31	and	and	CC
hansen-can-2020	139	32	then	then	RB
hansen-can-2020	139	33	to	to	TO
hansen-can-2020	139	34	look	look	VB
hansen-can-2020	139	35	even	even	RB
hansen-can-2020	139	36	closer	close	RBR
hansen-can-2020	139	37	and	and	CC
hansen-can-2020	139	38	make	make	VB
hansen-can-2020	139	39	sure	sure	JJ
hansen-can-2020	139	40	it	-PRON-	PRP
hansen-can-2020	139	41	is	be	VBZ
hansen-can-2020	139	42	the	the	DT
hansen-can-2020	139	43	type	type	NN
hansen-can-2020	139	44	of	of	IN
hansen-can-2020	139	45	nail	nail	NN
hansen-can-2020	139	46	a	a	DT
hansen-can-2020	139	47	machine	machine	NN
hansen-can-2020	139	48	-	-	HYPH
hansen-can-2020	139	49	learning	learn	VBG
hansen-can-2020	139	50	hammer	hammer	NN
hansen-can-2020	139	51	can	can	MD
hansen-can-2020	139	52	hit	hit	VB
hansen-can-2020	139	53	.	.	.
hansen-can-2020	140	1	Comment	comment	NN
hansen-can-2020	140	2	by	by	IN
hansen-can-2020	140	3	Mark	Mark	NNP
hansen-can-2020	140	4	Dehmlow	Dehmlow	NNP
hansen-can-2020	140	5	:	:	:
hansen-can-2020	140	6	Maybe	maybe	RB
hansen-can-2020	140	7	"	"	``
hansen-can-2020	140	8	think	think	VBP
hansen-can-2020	140	9	analytically	analytically	RB
hansen-can-2020	140	10	"	"	''
hansen-can-2020	140	11	?	?	.
hansen-can-2020	141	1	References	References	NNPS
hansen-can-2020	141	2	Barthel	Barthel	NNP
hansen-can-2020	141	3	,	,	,
hansen-can-2020	141	4	Simon	Simon	NNP
hansen-can-2020	141	5	,	,	,
hansen-can-2020	141	6	Sascha	Sascha	NNP
hansen-can-2020	141	7	Tönnies	Tönnies	NNP
hansen-can-2020	141	8	,	,	,
hansen-can-2020	141	9	and	and	CC
hansen-can-2020	141	10	Wolf	Wolf	NNP
hansen-can-2020	141	11	-	-	HYPH
hansen-can-2020	141	12	Tilo	Tilo	NNP
hansen-can-2020	141	13	Balke	Balke	NNP
hansen-can-2020	141	14	.	.	.
hansen-can-2020	142	1	2013	2013	CD
hansen-can-2020	142	2	.	.	.
hansen-can-2020	143	1	“	"	``
hansen-can-2020	143	2	Large	large	JJ
hansen-can-2020	143	3	-	-	HYPH
hansen-can-2020	143	4	Scale	scale	NN
hansen-can-2020	143	5	Experiments	experiment	NNS
hansen-can-2020	143	6	for	for	IN
hansen-can-2020	143	7	Mathematical	Mathematical	NNP
hansen-can-2020	143	8	Document	Document	NNP
hansen-can-2020	143	9	Classification	Classification	NNP
hansen-can-2020	143	10	.	.	.
hansen-can-2020	143	11	”	"	''
hansen-can-2020	143	12	In	in	IN
hansen-can-2020	143	13	Digital	Digital	NNP
hansen-can-2020	143	14	Libraries	Libraries	NNPS
hansen-can-2020	143	15	:	:	:
hansen-can-2020	143	16	Social	Social	NNP
hansen-can-2020	143	17	Media	Media	NNPS
hansen-can-2020	143	18	and	and	CC
hansen-can-2020	143	19	Community	Community	NNP
hansen-can-2020	143	20	Networks	Networks	NNP
hansen-can-2020	143	21	,	,	,
hansen-can-2020	143	22	edited	edit	VBN
hansen-can-2020	143	23	by	by	IN
hansen-can-2020	143	24	Shalini	Shalini	NNP
hansen-can-2020	143	25	R.	R.	NNP
hansen-can-2020	143	26	Urs	Urs	NNP
hansen-can-2020	143	27	,	,	,
hansen-can-2020	143	28	Jin	Jin	NNP
hansen-can-2020	143	29	-	-	HYPH
hansen-can-2020	143	30	Cheon	Cheon	NNP
hansen-can-2020	143	31	Na	Na	NNP
hansen-can-2020	143	32	,	,	,
hansen-can-2020	143	33	and	and	CC
hansen-can-2020	143	34	George	George	NNP
hansen-can-2020	143	35	Buchanan	Buchanan	NNP
hansen-can-2020	143	36	,	,	,
hansen-can-2020	143	37	83–92	83–92	NN
hansen-can-2020	143	38	.	.	.
hansen-can-2020	144	1	Cham	Cham	NNP
hansen-can-2020	144	2	:	:	:
hansen-can-2020	144	3	Springer	Springer	NNP
hansen-can-2020	144	4	International	International	NNP
hansen-can-2020	144	5	Publishing	Publishing	NNP
hansen-can-2020	144	6	.	.	.
hansen-can-2020	145	1	Clark	Clark	NNP
hansen-can-2020	145	2	,	,	,
hansen-can-2020	145	3	Jack	Jack	NNP
hansen-can-2020	145	4	.	.	.
hansen-can-2020	146	1	2015	2015	CD
hansen-can-2020	146	2	.	.	.
hansen-can-2020	147	1	“	"	``
hansen-can-2020	147	2	Why	why	WRB
hansen-can-2020	147	3	2015	2015	CD
hansen-can-2020	147	4	Was	be	VBD
hansen-can-2020	147	5	a	a	DT
hansen-can-2020	147	6	Breakthrough	Breakthrough	NNP
hansen-can-2020	147	7	Year	Year	NNP
hansen-can-2020	147	8	in	in	IN
hansen-can-2020	147	9	Artificial	Artificial	NNP
hansen-can-2020	147	10	Intelligence	Intelligence	NNP
hansen-can-2020	147	11	.	.	.
hansen-can-2020	147	12	”	"	''
hansen-can-2020	147	13	Bloomberg	Bloomberg	NNP
hansen-can-2020	147	14	,	,	,
hansen-can-2020	147	15	December	December	NNP
hansen-can-2020	147	16	8	8	CD
hansen-can-2020	147	17	,	,	,
hansen-can-2020	147	18	2015	2015	CD
hansen-can-2020	147	19	.	.	.
hansen-can-2020	147	20	https://www.bloomberg.com/news/articles/2015-12-08/why-2015-was-a-breakthrough-year-in-artificial-intelligence	https://www.bloomberg.com/news/articles/2015-12-08/why-2015-was-a-breakthrough-year-in-artificial-intelligence	NNP
hansen-can-2020	147	21	.	.	.
hansen-can-2020	148	1	Gandhi	Gandhi	NNP
hansen-can-2020	148	2	,	,	,
hansen-can-2020	148	3	Rohith	Rohith	NNP
hansen-can-2020	148	4	.	.	.
hansen-can-2020	149	1	2018	2018	CD
hansen-can-2020	149	2	.	.	.
hansen-can-2020	150	1	“	"	``
hansen-can-2020	150	2	Support	Support	NNP
hansen-can-2020	150	3	Vector	Vector	NNP
hansen-can-2020	150	4	Machine	Machine	NNP
hansen-can-2020	150	5	—	—	:
hansen-can-2020	150	6	Introduction	introduction	NN
hansen-can-2020	150	7	to	to	IN
hansen-can-2020	150	8	Machine	Machine	NNP
hansen-can-2020	150	9	Learning	Learning	NNP
hansen-can-2020	150	10	Algorithms	Algorithms	NNP
hansen-can-2020	150	11	.	.	.
hansen-can-2020	150	12	”	"	''
hansen-can-2020	150	13	Medium	medium	NN
hansen-can-2020	150	14	.	.	.
hansen-can-2020	151	1	July	July	NNP
hansen-can-2020	151	2	5	5	CD
hansen-can-2020	151	3	,	,	,
hansen-can-2020	151	4	2018	2018	CD
hansen-can-2020	151	5	.	.	.
hansen-can-2020	151	6	https://towardsdatascience.com/support-vector-machine-introduction-to-machine-learning-algorithms-934a444fca47	https://towardsdatascience.com/support-vector-machine-introduction-to-machine-learning-algorithms-934a444fca47	NNP
hansen-can-2020	151	7	.	.	.
hansen-can-2020	152	1	“	"	``
hansen-can-2020	152	2	GeoDeepDive	geodeepdive	NN
hansen-can-2020	152	3	:	:	:
hansen-can-2020	152	4	Project	Project	NNP
hansen-can-2020	152	5	Overview	Overview	NNP
hansen-can-2020	152	6	.	.	.
hansen-can-2020	152	7	”	"	''
hansen-can-2020	152	8	n.d	n.d	NNP
hansen-can-2020	152	9	.	.	NNP
hansen-can-2020	152	10	Accessed	Accessed	NNP
hansen-can-2020	152	11	May	May	NNP
hansen-can-2020	152	12	7	7	CD
hansen-can-2020	152	13	,	,	,
hansen-can-2020	152	14	2018	2018	CD
hansen-can-2020	152	15	.	.	.
hansen-can-2020	152	16	https://geodeepdive.org/about.html	https://geodeepdive.org/about.html	ADD
hansen-can-2020	152	17	.	.	.
hansen-can-2020	153	1	Kelly	Kelly	NNP
hansen-can-2020	153	2	,	,	,
hansen-can-2020	153	3	Kevin	Kevin	NNP
hansen-can-2020	153	4	.	.	.
hansen-can-2020	154	1	2014	2014	CD
hansen-can-2020	154	2	.	.	.
hansen-can-2020	155	1	“	"	``
hansen-can-2020	155	2	The	the	DT
hansen-can-2020	155	3	Three	three	CD
hansen-can-2020	155	4	Breakthroughs	Breakthroughs	NNPS
hansen-can-2020	155	5	That	that	WDT
hansen-can-2020	155	6	Have	have	VBP
hansen-can-2020	155	7	Finally	finally	RB
hansen-can-2020	155	8	Unleashed	unleash	VBN
hansen-can-2020	155	9	AI	AI	NNP
hansen-can-2020	155	10	on	on	IN
hansen-can-2020	155	11	the	the	DT
hansen-can-2020	155	12	World	World	NNP
hansen-can-2020	155	13	.	.	.
hansen-can-2020	155	14	”	"	''
hansen-can-2020	155	15	Wired	wire	VBN
hansen-can-2020	155	16	,	,	,
hansen-can-2020	155	17	October	October	NNP
hansen-can-2020	155	18	27	27	CD
hansen-can-2020	155	19	,	,	,
hansen-can-2020	155	20	2014	2014	CD
hansen-can-2020	155	21	.	.	.
hansen-can-2020	155	22	https://www.wired.com/2014/10/future-of-artificial-intelligence/.	https://www.wired.com/2014/10/future-of-artificial-intelligence/.	ADD
hansen-can-2020	156	1	“	"	``
hansen-can-2020	156	2	Microsoft	Microsoft	NNP
hansen-can-2020	156	3	Academic	Academic	NNP
hansen-can-2020	156	4	Increases	Increases	NNPS
hansen-can-2020	156	5	Power	Power	NNP
hansen-can-2020	156	6	of	of	IN
hansen-can-2020	156	7	Semantic	Semantic	NNP
hansen-can-2020	156	8	Search	Search	NNP
hansen-can-2020	156	9	by	by	IN
hansen-can-2020	156	10	Adding	add	VBG
hansen-can-2020	156	11	More	More	JJR
hansen-can-2020	156	12	Fields	field	NNS
hansen-can-2020	156	13	of	of	IN
hansen-can-2020	156	14	Study	Study	NNP
hansen-can-2020	156	15	.	.	.
hansen-can-2020	156	16	”	"	''
hansen-can-2020	156	17	2018	2018	CD
hansen-can-2020	156	18	.	.	.
hansen-can-2020	157	1	Microsoft	Microsoft	NNP
hansen-can-2020	157	2	Academic	Academic	NNP
hansen-can-2020	157	3	(	(	-LRB-
hansen-can-2020	157	4	blog	blog	NN
hansen-can-2020	157	5	)	)	-RRB-
hansen-can-2020	157	6	.	.	.
hansen-can-2020	158	1	February	February	NNP
hansen-can-2020	158	2	15	15	CD
hansen-can-2020	158	3	,	,	,
hansen-can-2020	158	4	2018	2018	CD
hansen-can-2020	158	5	.	.	.
hansen-can-2020	158	6	https://www.microsoft.com/en-us/research/project/academic/articles/microsoft-academic-increases-power-semantic-search-adding-fields-study/.	https://www.microsoft.com/en-us/research/project/academic/articles/microsoft-academic-increases-power-semantic-search-adding-fields-study/.	NNP
hansen-can-2020	159	1	Mikolov	Mikolov	NNP
hansen-can-2020	159	2	,	,	,
hansen-can-2020	159	3	Tomas	Tomas	NNP
hansen-can-2020	159	4	,	,	,
hansen-can-2020	159	5	Ilya	Ilya	NNP
hansen-can-2020	159	6	Sutskever	Sutskever	NNP
hansen-can-2020	159	7	,	,	,
hansen-can-2020	159	8	Kai	Kai	NNP
hansen-can-2020	159	9	Chen	Chen	NNP
hansen-can-2020	159	10	,	,	,
hansen-can-2020	159	11	Greg	Greg	NNP
hansen-can-2020	159	12	S	S	NNP
hansen-can-2020	159	13	Corrado	Corrado	NNP
hansen-can-2020	159	14	,	,	,
hansen-can-2020	159	15	and	and	CC
hansen-can-2020	159	16	Jeff	Jeff	NNP
hansen-can-2020	159	17	Dean	Dean	NNP
hansen-can-2020	159	18	.	.	.
hansen-can-2020	160	1	2013	2013	CD
hansen-can-2020	160	2	.	.	.
hansen-can-2020	161	1	“	"	``
hansen-can-2020	161	2	Distributed	distribute	VBN
hansen-can-2020	161	3	Representations	Representations	NNPS
hansen-can-2020	161	4	of	of	IN
hansen-can-2020	161	5	Words	Words	NNPS
hansen-can-2020	161	6	and	and	CC
hansen-can-2020	161	7	Phrases	Phrases	NNPS
hansen-can-2020	161	8	and	and	CC
hansen-can-2020	161	9	Their	-PRON-	PRP$
hansen-can-2020	161	10	Compositionality	compositionality	NN
hansen-can-2020	161	11	.	.	.
hansen-can-2020	161	12	”	"	''
hansen-can-2020	161	13	In	in	IN
hansen-can-2020	161	14	Advances	advance	NNS
hansen-can-2020	161	15	in	in	IN
hansen-can-2020	161	16	Neural	Neural	NNP
hansen-can-2020	161	17	Information	Information	NNP
hansen-can-2020	161	18	Processing	Processing	NNP
hansen-can-2020	161	19	Systems	Systems	NNP
hansen-can-2020	161	20	26	26	CD
hansen-can-2020	161	21	,	,	,
hansen-can-2020	161	22	edited	edit	VBN
hansen-can-2020	161	23	by	by	IN
hansen-can-2020	161	24	C.	C.	NNP
hansen-can-2020	161	25	J.	J.	NNP
hansen-can-2020	161	26	C.	C.	NNP
hansen-can-2020	161	27	Burges	Burges	NNP
hansen-can-2020	161	28	,	,	,
hansen-can-2020	161	29	L.	L.	NNP
hansen-can-2020	161	30	Bottou	Bottou	NNP
hansen-can-2020	161	31	,	,	,
hansen-can-2020	161	32	M.	M.	NNP
hansen-can-2020	161	33	Welling	Welling	NNP
hansen-can-2020	161	34	,	,	,
hansen-can-2020	161	35	Z.	Z.	NNP
hansen-can-2020	161	36	Ghahramani	Ghahramani	NNP
hansen-can-2020	161	37	,	,	,
hansen-can-2020	161	38	and	and	CC
hansen-can-2020	161	39	K.	K.	NNP
hansen-can-2020	161	40	Q.	Q.	NNP
hansen-can-2020	161	41	Weinberger	Weinberger	NNP
hansen-can-2020	161	42	,	,	,
hansen-can-2020	161	43	3111–3119	3111–3119	CD
hansen-can-2020	161	44	.	.	.
hansen-can-2020	162	1	Curran	Curran	NNP
hansen-can-2020	162	2	Associates	Associates	NNPS
hansen-can-2020	162	3	,	,	,
hansen-can-2020	162	4	Inc.	Inc.	NNP
hansen-can-2020	162	5	http://papers.nips.cc/paper/5021-distributed-representations-of-words-and-phrases-and-their-compositionality.pdf	http://papers.nips.cc/paper/5021-distributed-representations-of-words-and-phrases-and-their-compositionality.pdf	NNP
hansen-can-2020	162	6	.	.	.
hansen-can-2020	163	1	Parloff	Parloff	NNP
hansen-can-2020	163	2	,	,	,
hansen-can-2020	163	3	Roger	Roger	NNP
hansen-can-2020	163	4	.	.	.
hansen-can-2020	164	1	2016	2016	CD
hansen-can-2020	164	2	.	.	.
hansen-can-2020	165	1	“	"	``
hansen-can-2020	165	2	From	from	IN
hansen-can-2020	165	3	2016	2016	CD
hansen-can-2020	165	4	:	:	:
hansen-can-2020	165	5	Why	why	WRB
hansen-can-2020	165	6	Deep	deep	JJ
hansen-can-2020	165	7	Learning	learning	NN
hansen-can-2020	165	8	Is	be	VBZ
hansen-can-2020	165	9	Suddenly	suddenly	RB
hansen-can-2020	165	10	Changing	change	VBG
hansen-can-2020	165	11	Your	-PRON-	PRP$
hansen-can-2020	165	12	Life	life	NN
hansen-can-2020	165	13	.	.	.
hansen-can-2020	165	14	”	"	''
hansen-can-2020	165	15	Fortune	Fortune	NNP
hansen-can-2020	165	16	.	.	.
hansen-can-2020	166	1	September	September	NNP
hansen-can-2020	166	2	28	28	CD
hansen-can-2020	166	3	,	,	,
hansen-can-2020	166	4	2016	2016	CD
hansen-can-2020	166	5	.	.	.
hansen-can-2020	166	6	https://fortune.com/longform/ai-artificial-intelligence-deep-machine-learning/.	https://fortune.com/longform/ai-artificial-intelligence-deep-machine-learning/.	NNP
hansen-can-2020	167	1	Rahimi	Rahimi	NNP
hansen-can-2020	167	2	,	,	,
hansen-can-2020	167	3	Ali	Ali	NNP
hansen-can-2020	167	4	,	,	,
hansen-can-2020	167	5	and	and	CC
hansen-can-2020	167	6	Benjamin	Benjamin	NNP
hansen-can-2020	167	7	Recht	Recht	NNP
hansen-can-2020	167	8	.	.	.
hansen-can-2020	168	1	2017	2017	CD
hansen-can-2020	168	2	.	.	.
hansen-can-2020	169	1	“	"	``
hansen-can-2020	169	2	Back	back	RB
hansen-can-2020	169	3	When	when	WRB
hansen-can-2020	169	4	We	-PRON-	PRP
hansen-can-2020	169	5	Were	be	VBD
hansen-can-2020	169	6	Kids	kid	NNS
hansen-can-2020	169	7	.	.	.
hansen-can-2020	169	8	”	"	''
hansen-can-2020	169	9	Presentation	presentation	NN
hansen-can-2020	169	10	at	at	IN
hansen-can-2020	169	11	the	the	DT
hansen-can-2020	169	12	NIPS	NIPS	NNP
hansen-can-2020	169	13	2017	2017	CD
hansen-can-2020	169	14	Conference	Conference	NNP
hansen-can-2020	169	15	.	.	.
hansen-can-2020	170	1	https://www.youtube.com/watch?v=Qi1Yry33TQE	https://www.youtube.com/watch?v=Qi1Yry33TQE	NNP
hansen-can-2020	170	2	.	.	.
hansen-can-2020	171	1	Řehůřek	Řehůřek	NNP
hansen-can-2020	171	2	,	,	,
hansen-can-2020	171	3	Radim	Radim	NNP
hansen-can-2020	171	4	,	,	,
hansen-can-2020	171	5	and	and	CC
hansen-can-2020	171	6	Petr	Petr	NNP
hansen-can-2020	171	7	Sojka	sojka	RB
hansen-can-2020	171	8	.	.	.
hansen-can-2020	172	1	2008	2008	CD
hansen-can-2020	172	2	.	.	.
hansen-can-2020	173	1	“	"	``
hansen-can-2020	173	2	Automated	Automated	NNP
hansen-can-2020	173	3	Classification	Classification	NNP
hansen-can-2020	173	4	and	and	CC
hansen-can-2020	173	5	Categorization	Categorization	NNP
hansen-can-2020	173	6	of	of	IN
hansen-can-2020	173	7	 	  	_SP
hansen-can-2020	173	8	Mathematical	mathematical	JJ
hansen-can-2020	173	9	 	  	_SP
hansen-can-2020	173	10	Knowledge	knowledge	NN
hansen-can-2020	173	11	.	.	.
hansen-can-2020	173	12	”	"	''
hansen-can-2020	173	13	In	in	IN
hansen-can-2020	173	14	Intelligent	Intelligent	NNP
hansen-can-2020	173	15	Computer	Computer	NNP
hansen-can-2020	173	16	Mathematics	Mathematics	NNP
hansen-can-2020	173	17	,	,	,
hansen-can-2020	173	18	edited	edit	VBN
hansen-can-2020	173	19	by	by	IN
hansen-can-2020	173	20	Serge	Serge	NNP
hansen-can-2020	173	21	Autexier	Autexier	NNP
hansen-can-2020	173	22	,	,	,
hansen-can-2020	173	23	John	John	NNP
hansen-can-2020	173	24	Campbell	Campbell	NNP
hansen-can-2020	173	25	,	,	,
hansen-can-2020	173	26	Julio	Julio	NNP
hansen-can-2020	173	27	Rubio	Rubio	NNP
hansen-can-2020	173	28	,	,	,
hansen-can-2020	173	29	Volker	Volker	NNP
hansen-can-2020	173	30	Sorge	Sorge	NNP
hansen-can-2020	173	31	,	,	,
hansen-can-2020	173	32	Masakazu	Masakazu	NNP
hansen-can-2020	173	33	Suzuki	Suzuki	NNP
hansen-can-2020	173	34	,	,	,
hansen-can-2020	173	35	and	and	CC
hansen-can-2020	173	36	Freek	Freek	NNP
hansen-can-2020	173	37	Wiedijk	Wiedijk	NNP
hansen-can-2020	173	38	,	,	,
hansen-can-2020	173	39	543–57	543–57	CD
hansen-can-2020	173	40	.	.	.
hansen-can-2020	174	1	Berlin	Berlin	NNP
hansen-can-2020	174	2	:	:	:
hansen-can-2020	174	3	Springer	Springer	NNP
hansen-can-2020	174	4	-	-	HYPH
hansen-can-2020	174	5	Verlag	Verlag	NNP
hansen-can-2020	174	6	.	.	.
hansen-can-2020	175	1	Sangwani	Sangwani	NNP
hansen-can-2020	175	2	,	,	,
hansen-can-2020	175	3	Gaurav	Gaurav	NNP
hansen-can-2020	175	4	.	.	.
hansen-can-2020	176	1	2017	2017	CD
hansen-can-2020	176	2	.	.	.
hansen-can-2020	177	1	“	"	``
hansen-can-2020	177	2	2017	2017	CD
hansen-can-2020	177	3	Is	be	VBZ
hansen-can-2020	177	4	the	the	DT
hansen-can-2020	177	5	Year	Year	NNP
hansen-can-2020	177	6	of	of	IN
hansen-can-2020	177	7	Machine	Machine	NNP
hansen-can-2020	177	8	Learning	Learning	NNP
hansen-can-2020	177	9	.	.	.
hansen-can-2020	178	1	Here	here	RB
hansen-can-2020	178	2	’s	’	VBZ
hansen-can-2020	178	3	Why	why	WRB
hansen-can-2020	178	4	.	.	.
hansen-can-2020	178	5	”	"	''
hansen-can-2020	178	6	Business	Business	NNP
hansen-can-2020	178	7	Insider	Insider	NNP
hansen-can-2020	178	8	,	,	,
hansen-can-2020	178	9	January	January	NNP
hansen-can-2020	178	10	13	13	CD
hansen-can-2020	178	11	,	,	,
hansen-can-2020	178	12	2017	2017	CD
hansen-can-2020	178	13	.	.	.
hansen-can-2020	178	14	https://www.businessinsider.in/2017-is-the-year-of-machine-learning-heres-why/articleshow/56514535.cms	https://www.businessinsider.in/2017-is-the-year-of-machine-learning-heres-why/articleshow/56514535.cms	NNP
hansen-can-2020	178	15	.	.	.
hansen-can-2020	179	1	Shen	Shen	NNP
hansen-can-2020	179	2	,	,	,
hansen-can-2020	179	3	Zhihong	Zhihong	NNP
hansen-can-2020	179	4	,	,	,
hansen-can-2020	179	5	Hao	Hao	NNP
hansen-can-2020	179	6	Ma	Ma	NNP
hansen-can-2020	179	7	,	,	,
hansen-can-2020	179	8	and	and	CC
hansen-can-2020	179	9	Kuansan	Kuansan	NNP
hansen-can-2020	179	10	Wang	Wang	NNP
hansen-can-2020	179	11	.	.	.
hansen-can-2020	180	1	2018	2018	CD
hansen-can-2020	180	2	.	.	.
hansen-can-2020	181	1	“	"	``
hansen-can-2020	181	2	A	a	DT
hansen-can-2020	181	3	Web	web	NN
hansen-can-2020	181	4	-	-	HYPH
hansen-can-2020	181	5	Scale	scale	NN
hansen-can-2020	181	6	System	system	NN
hansen-can-2020	181	7	for	for	IN
hansen-can-2020	181	8	Scientific	Scientific	NNP
hansen-can-2020	181	9	Knowledge	Knowledge	NNP
hansen-can-2020	181	10	Exploration	Exploration	NNP
hansen-can-2020	181	11	.	.	.
hansen-can-2020	181	12	”	"	''
hansen-can-2020	181	13	Paper	paper	NN
hansen-can-2020	181	14	presented	present	VBD
hansen-can-2020	181	15	at	at	IN
hansen-can-2020	181	16	the	the	DT
hansen-can-2020	181	17	56th	56th	JJ
hansen-can-2020	181	18	Annual	Annual	NNP
hansen-can-2020	181	19	Meeting	Meeting	NNP
hansen-can-2020	181	20	of	of	IN
hansen-can-2020	181	21	the	the	DT
hansen-can-2020	181	22	Association	Association	NNP
hansen-can-2020	181	23	for	for	IN
hansen-can-2020	181	24	Computational	Computational	NNP
hansen-can-2020	181	25	Linguistics	Linguistics	NNPS
hansen-can-2020	181	26	,	,	,
hansen-can-2020	181	27	Melbourne	Melbourne	NNP
hansen-can-2020	181	28	,	,	,
hansen-can-2020	181	29	July	July	NNP
hansen-can-2020	181	30	2018	2018	CD
hansen-can-2020	181	31	.	.	.
hansen-can-2020	181	32	http://arxiv.org/abs/1805.12216	http://arxiv.org/abs/1805.12216	NNP
hansen-can-2020	181	33	.	.	.
hansen-can-2020	182	1	Tank	tank	NN
hansen-can-2020	182	2	,	,	,
hansen-can-2020	182	3	Aytekin	Aytekin	NNP
hansen-can-2020	182	4	.	.	.
hansen-can-2020	183	1	2017	2017	CD
hansen-can-2020	183	2	.	.	.
hansen-can-2020	184	1	“	"	``
hansen-can-2020	184	2	This	this	DT
hansen-can-2020	184	3	Is	be	VBZ
hansen-can-2020	184	4	the	the	DT
hansen-can-2020	184	5	Year	Year	NNP
hansen-can-2020	184	6	of	of	IN
hansen-can-2020	184	7	the	the	DT
hansen-can-2020	184	8	Machine	Machine	NNP
hansen-can-2020	184	9	Learning	Learning	NNP
hansen-can-2020	184	10	Revolution	Revolution	NNP
hansen-can-2020	184	11	.	.	.
hansen-can-2020	184	12	”	"	''
hansen-can-2020	184	13	Entrepreneur	entrepreneur	NN
hansen-can-2020	184	14	,	,	,
hansen-can-2020	184	15	January	January	NNP
hansen-can-2020	184	16	12	12	CD
hansen-can-2020	184	17	,	,	,
hansen-can-2020	184	18	2017	2017	CD
hansen-can-2020	184	19	.	.	.
hansen-can-2020	184	20	https://www.entrepreneur.com/article/287324	https://www.entrepreneur.com/article/287324	UH
hansen-can-2020	184	21	.	.	.
hansen-can-2020	185	1	Editor	editor	NN
hansen-can-2020	185	2	Notes	Notes	NNP
hansen-can-2020	185	3	There	there	EX
hansen-can-2020	185	4	are	be	VBP
hansen-can-2020	185	5	some	some	DT
hansen-can-2020	185	6	additional	additional	JJ
hansen-can-2020	185	7	stylistic	stylistic	JJ
hansen-can-2020	185	8	fixes	fix	NNS
hansen-can-2020	185	9	Mark	Mark	NNP
hansen-can-2020	185	10	and	and	CC
hansen-can-2020	185	11	I	-PRON-	PRP
hansen-can-2020	185	12	picked	pick	VBD
hansen-can-2020	185	13	up	up	RP
hansen-can-2020	185	14	on	on	IN
hansen-can-2020	185	15	,	,	,
hansen-can-2020	185	16	but	but	CC
hansen-can-2020	185	17	those	those	DT
hansen-can-2020	185	18	were	be	VBD
hansen-can-2020	185	19	easily	easily	RB
hansen-can-2020	185	20	made	make	VBN
hansen-can-2020	185	21	,	,	,
hansen-can-2020	185	22	and	and	CC
hansen-can-2020	185	23	there	there	EX
hansen-can-2020	185	24	's	be	VBZ
hansen-can-2020	185	25	only	only	RB
hansen-can-2020	185	26	a	a	DT
hansen-can-2020	185	27	handful	handful	NN
hansen-can-2020	185	28	of	of	IN
hansen-can-2020	185	29	very	very	RB
hansen-can-2020	185	30	small	small	JJ
hansen-can-2020	185	31	issues	issue	NNS
hansen-can-2020	185	32	for	for	IN
hansen-can-2020	185	33	Samuel	Samuel	NNP
hansen-can-2020	185	34	to	to	TO
hansen-can-2020	185	35	look	look	VB
hansen-can-2020	185	36	at	at	IN
hansen-can-2020	185	37	.	.	.
hansen-can-2020	186	1	The	the	DT
hansen-can-2020	186	2	author	author	NN
hansen-can-2020	186	3	addressed	address	VBD
hansen-can-2020	186	4	the	the	DT
hansen-can-2020	186	5	concerns	concern	NNS
hansen-can-2020	186	6	we	-PRON-	PRP
hansen-can-2020	186	7	had	have	VBD
hansen-can-2020	186	8	,	,	,
hansen-can-2020	186	9	especially	especially	RB
hansen-can-2020	186	10	about	about	IN
hansen-can-2020	186	11	the	the	DT
hansen-can-2020	186	12	conclusion	conclusion	NN
hansen-can-2020	186	13	,	,	,
hansen-can-2020	186	14	which	which	WDT
hansen-can-2020	186	15	is	be	VBZ
hansen-can-2020	186	16	nicely	nicely	RB
hansen-can-2020	186	17	moderated	moderate	VBN
hansen-can-2020	186	18	and	and	CC
hansen-can-2020	186	19	melds	meld	VBZ
hansen-can-2020	186	20	together	together	RB
hansen-can-2020	186	21	the	the	DT
hansen-can-2020	186	22	implications	implication	NNS
hansen-can-2020	186	23	of	of	IN
hansen-can-2020	186	24	the	the	DT
hansen-can-2020	186	25	strengths	strength	NNS
hansen-can-2020	186	26	/	/	SYM
hansen-can-2020	186	27	weaknesses	weakness	NNS
hansen-can-2020	186	28	of	of	IN
hansen-can-2020	186	29	ML	ML	NNP
hansen-can-2020	186	30	classifying	classify	VBG
hansen-can-2020	186	31	articulated	articulate	VBN
hansen-can-2020	186	32	earlier	early	RBR
hansen-can-2020	186	33	in	in	IN
hansen-can-2020	186	34	the	the	DT
hansen-can-2020	186	35	article	article	NN
hansen-can-2020	186	36	.	.	.