id	sid	tid	token	lemma	pos
zkie-zero-2020	1	1	From	from	IN
zkie-zero-2020	1	2	Zero	Zero	NNP
zkie-zero-2020	1	3	to	to	IN
zkie-zero-2020	1	4	Hero	Hero	NNP
zkie-zero-2020	1	5	:	:	:
zkie-zero-2020	1	6	Human	human	NN
zkie-zero-2020	1	7	-	-	HYPH
zkie-zero-2020	1	8	In	in	IN
zkie-zero-2020	1	9	-	-	HYPH
zkie-zero-2020	1	10	The	the	DT
zkie-zero-2020	1	11	-	-	HYPH
zkie-zero-2020	1	12	Loop	Loop	NNP
zkie-zero-2020	1	13	Entity	Entity	NNP
zkie-zero-2020	1	14	Linking	Linking	NNP
zkie-zero-2020	1	15	in	in	IN
zkie-zero-2020	1	16	Low	Low	NNP
zkie-zero-2020	1	17	Resource	Resource	NNP
zkie-zero-2020	1	18	Domains	Domains	NNP
zkie-zero-2020	1	19	Proceedings	Proceedings	NNPS
zkie-zero-2020	1	20	of	of	IN
zkie-zero-2020	1	21	the	the	DT
zkie-zero-2020	1	22	58th	58th	JJ
zkie-zero-2020	1	23	Annual	Annual	NNP
zkie-zero-2020	1	24	Meeting	Meeting	NNP
zkie-zero-2020	1	25	of	of	IN
zkie-zero-2020	1	26	the	the	DT
zkie-zero-2020	1	27	Association	Association	NNP
zkie-zero-2020	1	28	for	for	IN
zkie-zero-2020	1	29	Computational	Computational	NNP
zkie-zero-2020	1	30	Linguistics	Linguistics	NNP
zkie-zero-2020	1	31	,	,	,
zkie-zero-2020	1	32	pages	page	VBZ
zkie-zero-2020	1	33	6982–6993	6982–6993	CD
zkie-zero-2020	1	34	July	July	NNP
zkie-zero-2020	1	35	5	5	CD
zkie-zero-2020	1	36	-	-	SYM
zkie-zero-2020	1	37	10	10	CD
zkie-zero-2020	1	38	,	,	,
zkie-zero-2020	1	39	2020	2020	CD
zkie-zero-2020	1	40	.	.	.
zkie-zero-2020	1	41	c	c	NNP
zkie-zero-2020	1	42	©	©	NNP
zkie-zero-2020	1	43	2020	2020	CD
zkie-zero-2020	1	44	Association	Association	NNP
zkie-zero-2020	1	45	for	for	IN
zkie-zero-2020	1	46	Computational	Computational	NNP
zkie-zero-2020	1	47	Linguistics	Linguistics	NNP
zkie-zero-2020	1	48	6982	6982	CD
zkie-zero-2020	1	49	From	from	IN
zkie-zero-2020	1	50	Zero	Zero	NNP
zkie-zero-2020	1	51	to	to	IN
zkie-zero-2020	1	52	Hero	Hero	NNP
zkie-zero-2020	1	53	:	:	:
zkie-zero-2020	1	54	Human	human	NN
zkie-zero-2020	1	55	-	-	HYPH
zkie-zero-2020	1	56	In	in	IN
zkie-zero-2020	1	57	-	-	HYPH
zkie-zero-2020	1	58	The	the	DT
zkie-zero-2020	1	59	-	-	HYPH
zkie-zero-2020	1	60	Loop	Loop	NNP
zkie-zero-2020	1	61	Entity	Entity	NNP
zkie-zero-2020	1	62	Linking	Linking	NNP
zkie-zero-2020	1	63	in	in	IN
zkie-zero-2020	1	64	Low	Low	NNP
zkie-zero-2020	1	65	Resource	Resource	NNP
zkie-zero-2020	1	66	Domains	Domains	NNP
zkie-zero-2020	1	67	Jan	Jan	NNP
zkie-zero-2020	1	68	-	-	HYPH
zkie-zero-2020	1	69	Christoph	Christoph	NNP
zkie-zero-2020	1	70	Klie	Klie	NNP
zkie-zero-2020	1	71	Richard	Richard	NNP
zkie-zero-2020	1	72	Eckart	Eckart	NNP
zkie-zero-2020	1	73	de	de	NNP
zkie-zero-2020	1	74	Castilho	Castilho	NNP
zkie-zero-2020	1	75	Iryna	Iryna	NNP
zkie-zero-2020	1	76	Gurevych	Gurevych	NNP
zkie-zero-2020	1	77	Ubiquitous	ubiquitous	JJ
zkie-zero-2020	1	78	Knowledge	Knowledge	NNP
zkie-zero-2020	1	79	Processing	Processing	NNP
zkie-zero-2020	1	80	Lab	Lab	NNP
zkie-zero-2020	1	81	(	(	-LRB-
zkie-zero-2020	1	82	UKP	UKP	NNP
zkie-zero-2020	1	83	-	-	HYPH
zkie-zero-2020	1	84	TUDA	TUDA	NNP
zkie-zero-2020	1	85	)	)	-RRB-
zkie-zero-2020	1	86	Department	Department	NNP
zkie-zero-2020	1	87	of	of	IN
zkie-zero-2020	1	88	Computer	Computer	NNP
zkie-zero-2020	1	89	Science	Science	NNP
zkie-zero-2020	1	90	Technical	Technical	NNP
zkie-zero-2020	1	91	University	University	NNP
zkie-zero-2020	1	92	of	of	IN
zkie-zero-2020	1	93	Darmstadt	Darmstadt	NNP
zkie-zero-2020	1	94	,	,	,
zkie-zero-2020	1	95	Germany	Germany	NNP
zkie-zero-2020	1	96	www.ukp.tu-darmstadt.de	www.ukp.tu-darmstadt.de	JJ
zkie-zero-2020	1	97	Abstract	Abstract	NNP
zkie-zero-2020	1	98	Entity	Entity	NNP
zkie-zero-2020	1	99	linking	linking	NN
zkie-zero-2020	1	100	(	(	-LRB-
zkie-zero-2020	1	101	EL	el	NN
zkie-zero-2020	1	102	)	)	-RRB-
zkie-zero-2020	1	103	is	be	VBZ
zkie-zero-2020	1	104	concerned	concern	VBN
zkie-zero-2020	1	105	with	with	IN
zkie-zero-2020	1	106	disam-	disam-	JJ
zkie-zero-2020	1	107	biguating	biguating	NN
zkie-zero-2020	1	108	entity	entity	NN
zkie-zero-2020	1	109	mentions	mention	NNS
zkie-zero-2020	1	110	in	in	IN
zkie-zero-2020	1	111	a	a	DT
zkie-zero-2020	1	112	text	text	NN
zkie-zero-2020	1	113	against	against	IN
zkie-zero-2020	1	114	knowledge	knowledge	NN
zkie-zero-2020	1	115	bases	basis	NNS
zkie-zero-2020	1	116	(	(	-LRB-
zkie-zero-2020	1	117	KB	KB	NNP
zkie-zero-2020	1	118	)	)	-RRB-
zkie-zero-2020	1	119	.	.	.
zkie-zero-2020	2	1	It	-PRON-	PRP
zkie-zero-2020	2	2	is	be	VBZ
zkie-zero-2020	2	3	crucial	crucial	JJ
zkie-zero-2020	2	4	in	in	IN
zkie-zero-2020	2	5	a	a	DT
zkie-zero-2020	2	6	consid-	consid-	NN
zkie-zero-2020	2	7	erable	erable	JJ
zkie-zero-2020	2	8	number	number	NN
zkie-zero-2020	2	9	of	of	IN
zkie-zero-2020	2	10	fields	field	NNS
zkie-zero-2020	2	11	like	like	IN
zkie-zero-2020	2	12	humanities	humanity	NNS
zkie-zero-2020	2	13	,	,	,
zkie-zero-2020	2	14	tech-	tech-	JJ
zkie-zero-2020	2	15	nical	nical	JJ
zkie-zero-2020	2	16	writing	writing	NN
zkie-zero-2020	2	17	and	and	CC
zkie-zero-2020	2	18	biomedical	biomedical	JJ
zkie-zero-2020	2	19	sciences	science	NNS
zkie-zero-2020	2	20	to	to	TO
zkie-zero-2020	2	21	en-	en-	VB
zkie-zero-2020	2	22	rich	rich	JJ
zkie-zero-2020	2	23	texts	text	NNS
zkie-zero-2020	2	24	with	with	IN
zkie-zero-2020	2	25	semantics	semantic	NNS
zkie-zero-2020	2	26	and	and	CC
zkie-zero-2020	2	27	discover	discover	VBP
zkie-zero-2020	2	28	more	more	JJR
zkie-zero-2020	2	29	knowledge	knowledge	NN
zkie-zero-2020	2	30	.	.	.
zkie-zero-2020	3	1	The	the	DT
zkie-zero-2020	3	2	use	use	NN
zkie-zero-2020	3	3	of	of	IN
zkie-zero-2020	3	4	EL	el	NN
zkie-zero-2020	3	5	in	in	IN
zkie-zero-2020	3	6	such	such	JJ
zkie-zero-2020	3	7	domains	domain	NNS
zkie-zero-2020	3	8	requires	require	VBZ
zkie-zero-2020	3	9	handling	handle	VBG
zkie-zero-2020	3	10	noisy	noisy	JJ
zkie-zero-2020	3	11	texts	text	NNS
zkie-zero-2020	3	12	,	,	,
zkie-zero-2020	3	13	low	low	JJ
zkie-zero-2020	3	14	resource	resource	NN
zkie-zero-2020	3	15	set-	set-	NN
zkie-zero-2020	3	16	tings	ting	NNS
zkie-zero-2020	3	17	and	and	CC
zkie-zero-2020	3	18	domain	domain	NN
zkie-zero-2020	3	19	-	-	HYPH
zkie-zero-2020	3	20	specific	specific	JJ
zkie-zero-2020	3	21	KBs	kb	NNS
zkie-zero-2020	3	22	.	.	.
zkie-zero-2020	4	1	Existing	exist	VBG
zkie-zero-2020	4	2	ap-	ap-	JJ
zkie-zero-2020	4	3	proaches	proache	NNS
zkie-zero-2020	4	4	are	be	VBP
zkie-zero-2020	4	5	mostly	mostly	RB
zkie-zero-2020	4	6	inappropriate	inappropriate	JJ
zkie-zero-2020	4	7	for	for	IN
zkie-zero-2020	4	8	this	this	DT
zkie-zero-2020	4	9	,	,	,
zkie-zero-2020	4	10	as	as	IN
zkie-zero-2020	4	11	they	-PRON-	PRP
zkie-zero-2020	4	12	depend	depend	VBP
zkie-zero-2020	4	13	on	on	IN
zkie-zero-2020	4	14	training	training	NN
zkie-zero-2020	4	15	data	datum	NNS
zkie-zero-2020	4	16	.	.	.
zkie-zero-2020	5	1	However	however	RB
zkie-zero-2020	5	2	,	,	,
zkie-zero-2020	5	3	in	in	IN
zkie-zero-2020	5	4	the	the	DT
zkie-zero-2020	5	5	above	above	JJ
zkie-zero-2020	5	6	scenario	scenario	NN
zkie-zero-2020	5	7	,	,	,
zkie-zero-2020	5	8	there	there	EX
zkie-zero-2020	5	9	exists	exist	VBZ
zkie-zero-2020	5	10	hardly	hardly	RB
zkie-zero-2020	5	11	annotated	annotate	VBN
zkie-zero-2020	5	12	data	datum	NNS
zkie-zero-2020	5	13	,	,	,
zkie-zero-2020	5	14	and	and	CC
zkie-zero-2020	5	15	it	-PRON-	PRP
zkie-zero-2020	5	16	needs	need	VBZ
zkie-zero-2020	5	17	to	to	TO
zkie-zero-2020	5	18	be	be	VB
zkie-zero-2020	5	19	created	create	VBN
zkie-zero-2020	5	20	from	from	IN
zkie-zero-2020	5	21	scratch	scratch	NN
zkie-zero-2020	5	22	.	.	.
zkie-zero-2020	6	1	We	-PRON-	PRP
zkie-zero-2020	6	2	therefore	therefore	RB
zkie-zero-2020	6	3	present	present	VBP
zkie-zero-2020	6	4	a	a	DT
zkie-zero-2020	6	5	novel	novel	JJ
zkie-zero-2020	6	6	domain	domain	NN
zkie-zero-2020	6	7	-	-	HYPH
zkie-zero-2020	6	8	agnostic	agnostic	JJ
zkie-zero-2020	6	9	Human	human	NN
zkie-zero-2020	6	10	-	-	HYPH
zkie-zero-2020	6	11	In	in	IN
zkie-zero-2020	6	12	-	-	HYPH
zkie-zero-2020	6	13	The	the	DT
zkie-zero-2020	6	14	-	-	HYPH
zkie-zero-2020	6	15	Loop	Loop	NNP
zkie-zero-2020	6	16	annotation	annotation	NN
zkie-zero-2020	6	17	approach	approach	NN
zkie-zero-2020	6	18	:	:	:
zkie-zero-2020	6	19	we	-PRON-	PRP
zkie-zero-2020	6	20	use	use	VBP
zkie-zero-2020	6	21	recommenders	recommender	NNS
zkie-zero-2020	6	22	that	that	WDT
zkie-zero-2020	6	23	suggest	suggest	VBP
zkie-zero-2020	6	24	potential	potential	JJ
zkie-zero-2020	6	25	con-	con-	NN
zkie-zero-2020	6	26	cepts	cept	NNS
zkie-zero-2020	6	27	and	and	CC
zkie-zero-2020	6	28	adaptive	adaptive	JJ
zkie-zero-2020	6	29	candidate	candidate	NN
zkie-zero-2020	6	30	ranking	rank	VBG
zkie-zero-2020	6	31	,	,	,
zkie-zero-2020	6	32	thereby	thereby	RB
zkie-zero-2020	6	33	speeding	speed	VBG
zkie-zero-2020	6	34	up	up	RP
zkie-zero-2020	6	35	the	the	DT
zkie-zero-2020	6	36	overall	overall	JJ
zkie-zero-2020	6	37	annotation	annotation	NN
zkie-zero-2020	6	38	process	process	NN
zkie-zero-2020	6	39	and	and	CC
zkie-zero-2020	6	40	making	make	VBG
zkie-zero-2020	6	41	it	-PRON-	PRP
zkie-zero-2020	6	42	less	less	RBR
zkie-zero-2020	6	43	tedious	tedious	JJ
zkie-zero-2020	6	44	for	for	IN
zkie-zero-2020	6	45	users	user	NNS
zkie-zero-2020	6	46	.	.	.
zkie-zero-2020	7	1	We	-PRON-	PRP
zkie-zero-2020	7	2	evaluate	evaluate	VBP
zkie-zero-2020	7	3	our	-PRON-	PRP$
zkie-zero-2020	7	4	ranking	ranking	JJ
zkie-zero-2020	7	5	approach	approach	NN
zkie-zero-2020	7	6	in	in	IN
zkie-zero-2020	7	7	a	a	DT
zkie-zero-2020	7	8	simulation	simulation	NN
zkie-zero-2020	7	9	on	on	IN
zkie-zero-2020	7	10	diffi-	diffi-	NN
zkie-zero-2020	7	11	cult	cult	NN
zkie-zero-2020	7	12	texts	text	NNS
zkie-zero-2020	7	13	and	and	CC
zkie-zero-2020	7	14	show	show	VBP
zkie-zero-2020	7	15	that	that	IN
zkie-zero-2020	7	16	it	-PRON-	PRP
zkie-zero-2020	7	17	greatly	greatly	RB
zkie-zero-2020	7	18	outperforms	outperform	VBZ
zkie-zero-2020	7	19	a	a	DT
zkie-zero-2020	7	20	strong	strong	JJ
zkie-zero-2020	7	21	baseline	baseline	NN
zkie-zero-2020	7	22	in	in	IN
zkie-zero-2020	7	23	ranking	rank	VBG
zkie-zero-2020	7	24	accuracy	accuracy	NN
zkie-zero-2020	7	25	.	.	.
zkie-zero-2020	8	1	In	in	IN
zkie-zero-2020	8	2	a	a	DT
zkie-zero-2020	8	3	user	user	NN
zkie-zero-2020	8	4	study	study	NN
zkie-zero-2020	8	5	,	,	,
zkie-zero-2020	8	6	the	the	DT
zkie-zero-2020	8	7	annotation	annotation	NN
zkie-zero-2020	8	8	speed	speed	NN
zkie-zero-2020	8	9	improves	improve	VBZ
zkie-zero-2020	8	10	by	by	IN
zkie-zero-2020	8	11	35	35	CD
zkie-zero-2020	8	12	%	%	NN
zkie-zero-2020	8	13	compared	compare	VBN
zkie-zero-2020	8	14	to	to	IN
zkie-zero-2020	8	15	annotating	annotate	VBG
zkie-zero-2020	8	16	without	without	IN
zkie-zero-2020	8	17	interactive	interactive	JJ
zkie-zero-2020	8	18	support	support	NN
zkie-zero-2020	8	19	;	;	:
zkie-zero-2020	8	20	users	user	NNS
zkie-zero-2020	8	21	report	report	VBP
zkie-zero-2020	8	22	that	that	IN
zkie-zero-2020	8	23	they	-PRON-	PRP
zkie-zero-2020	8	24	strongly	strongly	RB
zkie-zero-2020	8	25	prefer	prefer	VBP
zkie-zero-2020	8	26	our	-PRON-	PRP$
zkie-zero-2020	8	27	system	system	NN
zkie-zero-2020	8	28	.	.	.
zkie-zero-2020	9	1	An	an	DT
zkie-zero-2020	9	2	open	open	JJ
zkie-zero-2020	9	3	-	-	HYPH
zkie-zero-2020	9	4	source	source	NN
zkie-zero-2020	9	5	and	and	CC
zkie-zero-2020	9	6	ready	ready	JJ
zkie-zero-2020	9	7	-	-	HYPH
zkie-zero-2020	9	8	to	to	TO
zkie-zero-2020	9	9	-	-	HYPH
zkie-zero-2020	9	10	use	use	VB
zkie-zero-2020	9	11	implementation	implementation	NN
zkie-zero-2020	9	12	based	base	VBN
zkie-zero-2020	9	13	on	on	IN
zkie-zero-2020	9	14	the	the	DT
zkie-zero-2020	9	15	text	text	NN
zkie-zero-2020	9	16	annotation	annotation	NN
zkie-zero-2020	9	17	platform	platform	NN
zkie-zero-2020	9	18	INCEpTION1	INCEpTION1	NNP
zkie-zero-2020	9	19	is	be	VBZ
zkie-zero-2020	9	20	made	make	VBN
zkie-zero-2020	9	21	available2	available2	NN
zkie-zero-2020	9	22	.	.	.
zkie-zero-2020	10	1	1	1	LS
zkie-zero-2020	10	2	Introduction	introduction	NN
zkie-zero-2020	10	3	Entity	Entity	NNP
zkie-zero-2020	10	4	linking	linking	NN
zkie-zero-2020	10	5	(	(	-LRB-
zkie-zero-2020	10	6	EL	el	NN
zkie-zero-2020	10	7	)	)	-RRB-
zkie-zero-2020	10	8	describes	describe	VBZ
zkie-zero-2020	10	9	the	the	DT
zkie-zero-2020	10	10	task	task	NN
zkie-zero-2020	10	11	of	of	IN
zkie-zero-2020	10	12	disam-	disam-	NN
zkie-zero-2020	10	13	biguating	biguating	NN
zkie-zero-2020	10	14	entity	entity	NN
zkie-zero-2020	10	15	mentions	mention	NNS
zkie-zero-2020	10	16	in	in	IN
zkie-zero-2020	10	17	a	a	DT
zkie-zero-2020	10	18	text	text	NN
zkie-zero-2020	10	19	by	by	IN
zkie-zero-2020	10	20	linking	link	VBG
zkie-zero-2020	10	21	them	-PRON-	PRP
zkie-zero-2020	10	22	to	to	IN
zkie-zero-2020	10	23	a	a	DT
zkie-zero-2020	10	24	knowledge	knowledge	NN
zkie-zero-2020	10	25	base	base	NN
zkie-zero-2020	10	26	(	(	-LRB-
zkie-zero-2020	10	27	KB	KB	NNP
zkie-zero-2020	10	28	)	)	-RRB-
zkie-zero-2020	10	29	,	,	,
zkie-zero-2020	10	30	e.g.	e.g.	RB
zkie-zero-2020	11	1	the	the	DT
zkie-zero-2020	11	2	text	text	NN
zkie-zero-2020	11	3	span	span	NN
zkie-zero-2020	11	4	Earl	Earl	NNP
zkie-zero-2020	11	5	of	of	IN
zkie-zero-2020	11	6	Orrery	Orrery	NNP
zkie-zero-2020	11	7	can	can	MD
zkie-zero-2020	11	8	be	be	VB
zkie-zero-2020	11	9	linked	link	VBN
zkie-zero-2020	11	10	to	to	IN
zkie-zero-2020	11	11	the	the	DT
zkie-zero-2020	11	12	KB	KB	NNP
zkie-zero-2020	11	13	entry	entry	NN
zkie-zero-2020	11	14	John	John	NNP
zkie-zero-2020	11	15	Boyle	Boyle	NNP
zkie-zero-2020	11	16	,	,	,
zkie-zero-2020	11	17	5	5	CD
zkie-zero-2020	11	18	.	.	.
zkie-zero-2020	12	1	Earl	Earl	NNP
zkie-zero-2020	12	2	of	of	IN
zkie-zero-2020	12	3	Cork	Cork	NNP
zkie-zero-2020	12	4	,	,	,
zkie-zero-2020	12	5	thereby	thereby	RB
zkie-zero-2020	12	6	disambiguating	disambiguate	VBG
zkie-zero-2020	12	7	it	-PRON-	PRP
zkie-zero-2020	12	8	.	.	.
zkie-zero-2020	13	1	EL	el	NN
zkie-zero-2020	13	2	is	be	VBZ
zkie-zero-2020	13	3	highly	highly	RB
zkie-zero-2020	13	4	beneficial	beneficial	JJ
zkie-zero-2020	13	5	in	in	IN
zkie-zero-2020	13	6	many	many	JJ
zkie-zero-2020	13	7	fields	field	NNS
zkie-zero-2020	13	8	like	like	IN
zkie-zero-2020	13	9	digital	digital	NNP
zkie-zero-2020	13	10	hu-	hu-	NN
zkie-zero-2020	13	11	manities	manitie	NNS
zkie-zero-2020	13	12	,	,	,
zkie-zero-2020	13	13	classics	classic	NNS
zkie-zero-2020	13	14	,	,	,
zkie-zero-2020	13	15	technical	technical	JJ
zkie-zero-2020	13	16	writing	writing	NN
zkie-zero-2020	13	17	or	or	CC
zkie-zero-2020	13	18	biomedical	biomedical	JJ
zkie-zero-2020	13	19	sciences	science	NNS
zkie-zero-2020	13	20	for	for	IN
zkie-zero-2020	13	21	applications	application	NNS
zkie-zero-2020	13	22	like	like	IN
zkie-zero-2020	13	23	search	search	NN
zkie-zero-2020	13	24	(	(	-LRB-
zkie-zero-2020	13	25	Meij	Meij	NNP
zkie-zero-2020	13	26	et	et	NNP
zkie-zero-2020	13	27	al	al	NNP
zkie-zero-2020	13	28	.	.	NNP
zkie-zero-2020	13	29	,	,	,
zkie-zero-2020	13	30	1	1	CD
zkie-zero-2020	13	31	https://inception-project.github.io	https://inception-project.github.io	NNP
zkie-zero-2020	13	32	2	2	CD
zkie-zero-2020	13	33	https://github.com/UKPLab/	https://github.com/UKPLab/	NNP
zkie-zero-2020	13	34	acl2020-interactive	acl2020-interactive	JJ
zkie-zero-2020	13	35	-	-	HYPH
zkie-zero-2020	13	36	entity	entity	NN
zkie-zero-2020	13	37	-	-	HYPH
zkie-zero-2020	13	38	linking	link	VBG
zkie-zero-2020	13	39	Figure	figure	NN
zkie-zero-2020	13	40	1	1	CD
zkie-zero-2020	13	41	:	:	:
zkie-zero-2020	13	42	Difficult	difficult	JJ
zkie-zero-2020	13	43	entity	entity	NN
zkie-zero-2020	13	44	mentions	mention	NNS
zkie-zero-2020	13	45	with	with	IN
zkie-zero-2020	13	46	their	-PRON-	PRP$
zkie-zero-2020	13	47	linked	link	VBN
zkie-zero-2020	13	48	en-	en-	NN
zkie-zero-2020	13	49	tities	titie	NNS
zkie-zero-2020	13	50	:	:	:
zkie-zero-2020	13	51	1	1	LS
zkie-zero-2020	13	52	)	)	-RRB-
zkie-zero-2020	13	53	Name	name	NN
zkie-zero-2020	13	54	variations	variation	NNS
zkie-zero-2020	13	55	,	,	,
zkie-zero-2020	13	56	2	2	LS
zkie-zero-2020	13	57	)	)	-RRB-
zkie-zero-2020	13	58	Spelling	Spelling	NNP
zkie-zero-2020	13	59	Variation	Variation	NNP
zkie-zero-2020	13	60	,	,	,
zkie-zero-2020	13	61	3	3	CD
zkie-zero-2020	13	62	)	)	-RRB-
zkie-zero-2020	13	63	Am-	am-	CD
zkie-zero-2020	13	64	biguity	biguity	NN
zkie-zero-2020	13	65	2014	2014	CD
zkie-zero-2020	13	66	)	)	-RRB-
zkie-zero-2020	13	67	,	,	,
zkie-zero-2020	13	68	semantic	semantic	JJ
zkie-zero-2020	13	69	enrichment	enrichment	NN
zkie-zero-2020	13	70	(	(	-LRB-
zkie-zero-2020	13	71	Schlögl	Schlögl	NNP
zkie-zero-2020	13	72	and	and	CC
zkie-zero-2020	13	73	Lejtovicz	Lejtovicz	NNP
zkie-zero-2020	13	74	,	,	,
zkie-zero-2020	13	75	2017	2017	CD
zkie-zero-2020	13	76	)	)	-RRB-
zkie-zero-2020	13	77	or	or	CC
zkie-zero-2020	13	78	information	information	NN
zkie-zero-2020	13	79	extraction	extraction	NN
zkie-zero-2020	13	80	(	(	-LRB-
zkie-zero-2020	13	81	Nooralahzadeh	Nooralahzadeh	NNP
zkie-zero-2020	13	82	and	and	CC
zkie-zero-2020	13	83	Øvrelid	Øvrelid	NNP
zkie-zero-2020	13	84	,	,	,
zkie-zero-2020	13	85	2018	2018	CD
zkie-zero-2020	13	86	)	)	-RRB-
zkie-zero-2020	13	87	.	.	.
zkie-zero-2020	14	1	These	these	DT
zkie-zero-2020	14	2	are	be	VBP
zkie-zero-2020	14	3	overwhelmingly	overwhelmingly	RB
zkie-zero-2020	14	4	low	low	JJ
zkie-zero-2020	14	5	-	-	HYPH
zkie-zero-2020	14	6	resource	resource	NN
zkie-zero-2020	14	7	settings	setting	NNS
zkie-zero-2020	14	8	:	:	:
zkie-zero-2020	14	9	often	often	RB
zkie-zero-2020	14	10	,	,	,
zkie-zero-2020	14	11	no	no	DT
zkie-zero-2020	14	12	data	datum	NNS
zkie-zero-2020	14	13	annotated	annotate	VBD
zkie-zero-2020	14	14	ex-	ex-	RB
zkie-zero-2020	14	15	ists	ist	NNS
zkie-zero-2020	14	16	;	;	:
zkie-zero-2020	14	17	coverage	coverage	NN
zkie-zero-2020	14	18	of	of	IN
zkie-zero-2020	14	19	open	open	JJ
zkie-zero-2020	14	20	-	-	HYPH
zkie-zero-2020	14	21	domain	domain	NN
zkie-zero-2020	14	22	knowledge	knowledge	NN
zkie-zero-2020	14	23	bases	basis	NNS
zkie-zero-2020	14	24	like	like	IN
zkie-zero-2020	14	25	Wikipedia	Wikipedia	NNP
zkie-zero-2020	14	26	or	or	CC
zkie-zero-2020	14	27	DBPedia	DBPedia	NNP
zkie-zero-2020	14	28	is	be	VBZ
zkie-zero-2020	14	29	low	low	JJ
zkie-zero-2020	14	30	.	.	.
zkie-zero-2020	15	1	Therefore	therefore	RB
zkie-zero-2020	15	2	,	,	,
zkie-zero-2020	15	3	en-	en-	CC
zkie-zero-2020	15	4	tity	tity	NN
zkie-zero-2020	15	5	linking	linking	NN
zkie-zero-2020	15	6	is	be	VBZ
zkie-zero-2020	15	7	frequently	frequently	RB
zkie-zero-2020	15	8	performed	perform	VBN
zkie-zero-2020	15	9	against	against	IN
zkie-zero-2020	15	10	domain-	domain-	JJ
zkie-zero-2020	15	11	specific	specific	JJ
zkie-zero-2020	15	12	knowledge	knowledge	NN
zkie-zero-2020	15	13	bases	basis	NNS
zkie-zero-2020	15	14	(	(	-LRB-
zkie-zero-2020	15	15	Munnelly	munnelly	RB
zkie-zero-2020	15	16	and	and	CC
zkie-zero-2020	15	17	Lawless	Lawless	NNP
zkie-zero-2020	15	18	,	,	,
zkie-zero-2020	15	19	2018a	2018a	CD
zkie-zero-2020	15	20	;	;	:
zkie-zero-2020	15	21	Bartsch	Bartsch	NNP
zkie-zero-2020	15	22	,	,	,
zkie-zero-2020	15	23	2004	2004	CD
zkie-zero-2020	15	24	)	)	-RRB-
zkie-zero-2020	15	25	.	.	.
zkie-zero-2020	16	1	In	in	IN
zkie-zero-2020	16	2	these	these	DT
zkie-zero-2020	16	3	scenarios	scenario	NNS
zkie-zero-2020	16	4	,	,	,
zkie-zero-2020	16	5	the	the	DT
zkie-zero-2020	16	6	first	first	JJ
zkie-zero-2020	16	7	crucial	crucial	JJ
zkie-zero-2020	16	8	step	step	NN
zkie-zero-2020	16	9	is	be	VBZ
zkie-zero-2020	16	10	to	to	TO
zkie-zero-2020	16	11	ob-	ob-	XX
zkie-zero-2020	16	12	tain	tain	VB
zkie-zero-2020	16	13	annotated	annotate	VBN
zkie-zero-2020	16	14	data	datum	NNS
zkie-zero-2020	16	15	.	.	.
zkie-zero-2020	17	1	This	this	DT
zkie-zero-2020	17	2	data	data	NN
zkie-zero-2020	17	3	can	can	MD
zkie-zero-2020	17	4	then	then	RB
zkie-zero-2020	17	5	be	be	VB
zkie-zero-2020	17	6	either	either	CC
zkie-zero-2020	17	7	directly	directly	RB
zkie-zero-2020	17	8	used	use	VBN
zkie-zero-2020	17	9	by	by	IN
zkie-zero-2020	17	10	researchers	researcher	NNS
zkie-zero-2020	17	11	for	for	IN
zkie-zero-2020	17	12	their	-PRON-	PRP$
zkie-zero-2020	17	13	downstream	downstream	JJ
zkie-zero-2020	17	14	task	task	NN
zkie-zero-2020	17	15	or	or	CC
zkie-zero-2020	17	16	to	to	TO
zkie-zero-2020	17	17	train	train	VB
zkie-zero-2020	17	18	machine	machine	NN
zkie-zero-2020	17	19	learning	learning	NN
zkie-zero-2020	17	20	models	model	NNS
zkie-zero-2020	17	21	for	for	IN
zkie-zero-2020	17	22	au-	au-	CC
zkie-zero-2020	17	23	tomatic	tomatic	JJ
zkie-zero-2020	17	24	annotation	annotation	NN
zkie-zero-2020	17	25	.	.	.
zkie-zero-2020	18	1	For	for	IN
zkie-zero-2020	18	2	this	this	DT
zkie-zero-2020	18	3	initial	initial	JJ
zkie-zero-2020	18	4	data	data	NN
zkie-zero-2020	18	5	creation	creation	NN
zkie-zero-2020	18	6	step	step	NN
zkie-zero-2020	18	7	,	,	,
zkie-zero-2020	18	8	we	-PRON-	PRP
zkie-zero-2020	18	9	developed	develop	VBD
zkie-zero-2020	18	10	a	a	DT
zkie-zero-2020	18	11	novel	novel	JJ
zkie-zero-2020	18	12	Human	Human	NNP
zkie-zero-2020	18	13	-	-	HYPH
zkie-zero-2020	18	14	In	in	IN
zkie-zero-2020	18	15	-	-	HYPH
zkie-zero-2020	18	16	The	The	NNP
zkie-zero-2020	18	17	-	-	HYPH
zkie-zero-2020	18	18	Loop	Loop	NNP
zkie-zero-2020	18	19	(	(	-LRB-
zkie-zero-2020	18	20	HITL	HITL	NNP
zkie-zero-2020	18	21	)	)	-RRB-
zkie-zero-2020	18	22	annotation	annotation	NN
zkie-zero-2020	18	23	approach	approach	NN
zkie-zero-2020	18	24	.	.	.
zkie-zero-2020	19	1	Manual	manual	JJ
zkie-zero-2020	19	2	annotation	annotation	NN
zkie-zero-2020	19	3	is	be	VBZ
zkie-zero-2020	19	4	laborious	laborious	JJ
zkie-zero-2020	19	5	and	and	CC
zkie-zero-2020	19	6	often	often	RB
zkie-zero-2020	19	7	prohibitively	prohibitively	RB
zkie-zero-2020	19	8	expensive	expensive	JJ
zkie-zero-2020	19	9	.	.	.
zkie-zero-2020	20	1	To	to	TO
zkie-zero-2020	20	2	improve	improve	VB
zkie-zero-2020	20	3	annotation	annotation	NN
zkie-zero-2020	20	4	speed	speed	NN
zkie-zero-2020	20	5	and	and	CC
zkie-zero-2020	20	6	quality	quality	NN
zkie-zero-2020	20	7	,	,	,
zkie-zero-2020	20	8	we	-PRON-	PRP
zkie-zero-2020	20	9	there-	there-	VBP
zkie-zero-2020	20	10	fore	fore	NNP
zkie-zero-2020	20	11	add	add	VBP
zkie-zero-2020	20	12	interactive	interactive	JJ
zkie-zero-2020	20	13	machine	machine	NN
zkie-zero-2020	20	14	learning	learning	NN
zkie-zero-2020	20	15	annotation	annotation	NN
zkie-zero-2020	20	16	support	support	NN
zkie-zero-2020	20	17	that	that	WDT
zkie-zero-2020	20	18	helps	help	VBZ
zkie-zero-2020	20	19	the	the	DT
zkie-zero-2020	20	20	user	user	NN
zkie-zero-2020	20	21	find	find	VB
zkie-zero-2020	20	22	entities	entity	NNS
zkie-zero-2020	20	23	in	in	IN
zkie-zero-2020	20	24	the	the	DT
zkie-zero-2020	20	25	text	text	NN
zkie-zero-2020	20	26	and	and	CC
zkie-zero-2020	20	27	select	select	VB
zkie-zero-2020	20	28	the	the	DT
zkie-zero-2020	20	29	correct	correct	JJ
zkie-zero-2020	20	30	knowledge	knowledge	NN
zkie-zero-2020	20	31	base	base	NN
zkie-zero-2020	20	32	entries	entry	NNS
zkie-zero-2020	20	33	for	for	IN
zkie-zero-2020	20	34	them	-PRON-	PRP
zkie-zero-2020	20	35	.	.	.
zkie-zero-2020	21	1	The	the	DT
zkie-zero-2020	21	2	more	more	JJR
zkie-zero-2020	21	3	entities	entity	NNS
zkie-zero-2020	21	4	are	be	VBP
zkie-zero-2020	21	5	annotated	annotate	VBN
zkie-zero-2020	21	6	,	,	,
zkie-zero-2020	21	7	the	the	DT
zkie-zero-2020	21	8	better	well	JJR
zkie-zero-2020	21	9	the	the	DT
zkie-zero-2020	21	10	annotation	annotation	NN
zkie-zero-2020	21	11	support	support	NN
zkie-zero-2020	21	12	will	will	MD
zkie-zero-2020	21	13	be	be	VB
zkie-zero-2020	21	14	.	.	.
zkie-zero-2020	22	1	Throughout	throughout	IN
zkie-zero-2020	22	2	this	this	DT
zkie-zero-2020	22	3	work	work	NN
zkie-zero-2020	22	4	,	,	,
zkie-zero-2020	22	5	we	-PRON-	PRP
zkie-zero-2020	22	6	focus	focus	VBP
zkie-zero-2020	22	7	on	on	IN
zkie-zero-2020	22	8	texts	text	NNS
zkie-zero-2020	22	9	from	from	IN
zkie-zero-2020	22	10	digital	digital	JJ
zkie-zero-2020	22	11	humanities	humanity	NNS
zkie-zero-2020	22	12	,	,	,
zkie-zero-2020	22	13	to	to	TO
zkie-zero-2020	22	14	be	be	VB
zkie-zero-2020	22	15	more	more	RBR
zkie-zero-2020	22	16	precise	precise	JJ
zkie-zero-2020	22	17	,	,	,
zkie-zero-2020	22	18	texts	text	NNS
zkie-zero-2020	22	19	written	write	VBN
zkie-zero-2020	22	20	in	in	IN
zkie-zero-2020	22	21	Early	early	JJ
zkie-zero-2020	22	22	Modern	modern	JJ
zkie-zero-2020	22	23	English	English	NNP
zkie-zero-2020	22	24	texts	text	NNS
zkie-zero-2020	22	25	,	,	,
zkie-zero-2020	22	26	including	include	VBG
zkie-zero-2020	22	27	poems	poem	NNS
zkie-zero-2020	22	28	,	,	,
zkie-zero-2020	22	29	biographies	biography	NNS
zkie-zero-2020	22	30	,	,	,
zkie-zero-2020	22	31	novels	novel	NNS
zkie-zero-2020	22	32	as	as	RB
zkie-zero-2020	22	33	well	well	RB
zkie-zero-2020	22	34	as	as	IN
zkie-zero-2020	22	35	legal	legal	JJ
zkie-zero-2020	22	36	documents	document	NNS
zkie-zero-2020	22	37	.	.	.
zkie-zero-2020	23	1	In	in	IN
zkie-zero-2020	23	2	https://inception-project.github.io	https://inception-project.github.io	NNP
zkie-zero-2020	23	3	https://inception-project.github.io	https://inception-project.github.io	UH
zkie-zero-2020	23	4	https://github.com/UKPLab/acl2020-interactive-entity-linking	https://github.com/ukplab/acl2020-interactive-entity-linke	VBG
zkie-zero-2020	23	5	https://github.com/UKPLab/acl2020-interactive-entity-linking	https://github.com/ukplab/acl2020-interactive-entity-linke	VBG
zkie-zero-2020	23	6	6983	6983	CD
zkie-zero-2020	23	7	this	this	DT
zkie-zero-2020	23	8	domain	domain	NN
zkie-zero-2020	23	9	,	,	,
zkie-zero-2020	23	10	texts	text	NNS
zkie-zero-2020	23	11	are	be	VBP
zkie-zero-2020	23	12	noisy	noisy	JJ
zkie-zero-2020	23	13	as	as	IN
zkie-zero-2020	23	14	they	-PRON-	PRP
zkie-zero-2020	23	15	were	be	VBD
zkie-zero-2020	23	16	written	write	VBN
zkie-zero-2020	23	17	in	in	IN
zkie-zero-2020	23	18	times	time	NNS
zkie-zero-2020	23	19	where	where	WRB
zkie-zero-2020	23	20	orthography	orthography	NNP
zkie-zero-2020	23	21	was	be	VBD
zkie-zero-2020	23	22	rather	rather	RB
zkie-zero-2020	23	23	incidental	incidental	JJ
zkie-zero-2020	23	24	or	or	CC
zkie-zero-2020	23	25	due	due	IN
zkie-zero-2020	23	26	to	to	IN
zkie-zero-2020	23	27	OCR	ocr	NN
zkie-zero-2020	23	28	and	and	CC
zkie-zero-2020	23	29	transcription	transcription	NN
zkie-zero-2020	23	30	errors	error	NNS
zkie-zero-2020	23	31	(	(	-LRB-
zkie-zero-2020	23	32	see	see	VB
zkie-zero-2020	23	33	Fig	Fig	NNP
zkie-zero-2020	23	34	.	.	.
zkie-zero-2020	24	1	1	1	LS
zkie-zero-2020	24	2	)	)	-RRB-
zkie-zero-2020	24	3	.	.	.
zkie-zero-2020	25	1	Tools	tool	NNS
zkie-zero-2020	25	2	like	like	IN
zkie-zero-2020	25	3	named	name	VBN
zkie-zero-2020	25	4	entity	entity	NN
zkie-zero-2020	25	5	recognizers	recognizer	NNS
zkie-zero-2020	25	6	are	be	VBP
zkie-zero-2020	25	7	unavailable	unavailable	JJ
zkie-zero-2020	25	8	or	or	CC
zkie-zero-2020	25	9	perform	perform	VBP
zkie-zero-2020	25	10	poorly	poorly	RB
zkie-zero-2020	25	11	(	(	-LRB-
zkie-zero-2020	25	12	Erdmann	Erdmann	NNP
zkie-zero-2020	25	13	et	et	FW
zkie-zero-2020	25	14	al	al	NNP
zkie-zero-2020	25	15	.	.	NNP
zkie-zero-2020	25	16	,	,	,
zkie-zero-2020	25	17	2019	2019	CD
zkie-zero-2020	25	18	)	)	-RRB-
zkie-zero-2020	25	19	.	.	.
zkie-zero-2020	26	1	We	-PRON-	PRP
zkie-zero-2020	26	2	demonstrate	demonstrate	VBP
zkie-zero-2020	26	3	the	the	DT
zkie-zero-2020	26	4	effectiveness	effectiveness	NN
zkie-zero-2020	26	5	of	of	IN
zkie-zero-2020	26	6	our	-PRON-	PRP$
zkie-zero-2020	26	7	ap-	ap-	NNP
zkie-zero-2020	26	8	proach	proach	NN
zkie-zero-2020	26	9	with	with	IN
zkie-zero-2020	26	10	extensive	extensive	JJ
zkie-zero-2020	26	11	simulation	simulation	NN
zkie-zero-2020	26	12	as	as	RB
zkie-zero-2020	26	13	well	well	RB
zkie-zero-2020	26	14	as	as	IN
zkie-zero-2020	26	15	a	a	DT
zkie-zero-2020	26	16	user	user	NN
zkie-zero-2020	26	17	study	study	NN
zkie-zero-2020	26	18	on	on	IN
zkie-zero-2020	26	19	different	different	JJ
zkie-zero-2020	26	20	,	,	,
zkie-zero-2020	26	21	challenging	challenge	VBG
zkie-zero-2020	26	22	datasets	dataset	NNS
zkie-zero-2020	26	23	.	.	.
zkie-zero-2020	27	1	We	-PRON-	PRP
zkie-zero-2020	27	2	imple-	imple-	VBP
zkie-zero-2020	27	3	ment	ment	JJ
zkie-zero-2020	27	4	our	-PRON-	PRP$
zkie-zero-2020	27	5	approach	approach	NN
zkie-zero-2020	27	6	based	base	VBN
zkie-zero-2020	27	7	on	on	IN
zkie-zero-2020	27	8	the	the	DT
zkie-zero-2020	27	9	open	open	JJ
zkie-zero-2020	27	10	-	-	HYPH
zkie-zero-2020	27	11	source	source	NN
zkie-zero-2020	27	12	anno-	anno-	NN
zkie-zero-2020	27	13	tation	tation	NN
zkie-zero-2020	27	14	platform	platform	NN
zkie-zero-2020	27	15	INCEpTION	inception	CD
zkie-zero-2020	27	16	(	(	-LRB-
zkie-zero-2020	27	17	Klie	Klie	NNP
zkie-zero-2020	27	18	et	et	FW
zkie-zero-2020	27	19	al	al	NNP
zkie-zero-2020	27	20	.	.	NNP
zkie-zero-2020	27	21	,	,	,
zkie-zero-2020	27	22	2018	2018	CD
zkie-zero-2020	27	23	)	)	-RRB-
zkie-zero-2020	27	24	and	and	CC
zkie-zero-2020	27	25	publish	publish	VB
zkie-zero-2020	27	26	all	all	DT
zkie-zero-2020	27	27	datasets	dataset	NNS
zkie-zero-2020	27	28	and	and	CC
zkie-zero-2020	27	29	code	code	NN
zkie-zero-2020	27	30	.	.	.
zkie-zero-2020	28	1	Our	-PRON-	PRP$
zkie-zero-2020	28	2	contributions	contribution	NNS
zkie-zero-2020	28	3	are	be	VBP
zkie-zero-2020	28	4	the	the	DT
zkie-zero-2020	28	5	following	follow	VBG
zkie-zero-2020	28	6	:	:	:
zkie-zero-2020	28	7	1	1	CD
zkie-zero-2020	28	8	.	.	.
zkie-zero-2020	29	1	We	-PRON-	PRP
zkie-zero-2020	29	2	present	present	VBP
zkie-zero-2020	29	3	a	a	DT
zkie-zero-2020	29	4	generic	generic	JJ
zkie-zero-2020	29	5	,	,	,
zkie-zero-2020	29	6	KB	KB	NNP
zkie-zero-2020	29	7	-	-	HYPH
zkie-zero-2020	29	8	agnostic	agnostic	JJ
zkie-zero-2020	29	9	annotation	annotation	NN
zkie-zero-2020	29	10	approach	approach	NN
zkie-zero-2020	29	11	for	for	IN
zkie-zero-2020	29	12	low	low	JJ
zkie-zero-2020	29	13	-	-	HYPH
zkie-zero-2020	29	14	resource	resource	NN
zkie-zero-2020	29	15	settings	setting	NNS
zkie-zero-2020	29	16	and	and	CC
zkie-zero-2020	29	17	pro-	pro-	NN
zkie-zero-2020	29	18	vide	vide	VBP
zkie-zero-2020	29	19	a	a	DT
zkie-zero-2020	29	20	ready	ready	JJ
zkie-zero-2020	29	21	-	-	HYPH
zkie-zero-2020	29	22	to	to	TO
zkie-zero-2020	29	23	-	-	HYPH
zkie-zero-2020	29	24	use	use	VB
zkie-zero-2020	29	25	implementation	implementation	NN
zkie-zero-2020	29	26	so	so	IN
zkie-zero-2020	29	27	that	that	IN
zkie-zero-2020	29	28	researchers	researcher	NNS
zkie-zero-2020	29	29	can	can	MD
zkie-zero-2020	29	30	easily	easily	RB
zkie-zero-2020	29	31	annotate	annotate	VB
zkie-zero-2020	29	32	data	datum	NNS
zkie-zero-2020	29	33	for	for	IN
zkie-zero-2020	29	34	their	-PRON-	PRP$
zkie-zero-2020	29	35	use	use	NN
zkie-zero-2020	29	36	cases	case	NNS
zkie-zero-2020	29	37	.	.	.
zkie-zero-2020	30	1	We	-PRON-	PRP
zkie-zero-2020	30	2	validate	validate	VBP
zkie-zero-2020	30	3	our	-PRON-	PRP$
zkie-zero-2020	30	4	approach	approach	NN
zkie-zero-2020	30	5	exten-	exten-	NNP
zkie-zero-2020	30	6	sively	sively	RB
zkie-zero-2020	30	7	in	in	IN
zkie-zero-2020	30	8	a	a	DT
zkie-zero-2020	30	9	simulation	simulation	NN
zkie-zero-2020	30	10	and	and	CC
zkie-zero-2020	30	11	in	in	IN
zkie-zero-2020	30	12	a	a	DT
zkie-zero-2020	30	13	user	user	NN
zkie-zero-2020	30	14	study	study	NN
zkie-zero-2020	30	15	.	.	.
zkie-zero-2020	31	1	2	2	LS
zkie-zero-2020	31	2	.	.	.
zkie-zero-2020	32	1	We	-PRON-	PRP
zkie-zero-2020	32	2	show	show	VBP
zkie-zero-2020	32	3	that	that	IN
zkie-zero-2020	32	4	statistical	statistical	JJ
zkie-zero-2020	32	5	machine	machine	NN
zkie-zero-2020	32	6	learning	learning	NN
zkie-zero-2020	32	7	models	model	NNS
zkie-zero-2020	32	8	can	can	MD
zkie-zero-2020	32	9	be	be	VB
zkie-zero-2020	32	10	used	use	VBN
zkie-zero-2020	32	11	in	in	IN
zkie-zero-2020	32	12	an	an	DT
zkie-zero-2020	32	13	interactive	interactive	JJ
zkie-zero-2020	32	14	entity	entity	NN
zkie-zero-2020	32	15	linking	link	VBG
zkie-zero-2020	32	16	setting	set	VBG
zkie-zero-2020	32	17	to	to	TO
zkie-zero-2020	32	18	improve	improve	VB
zkie-zero-2020	32	19	annotation	annotation	NN
zkie-zero-2020	32	20	speed	speed	NN
zkie-zero-2020	32	21	by	by	IN
zkie-zero-2020	32	22	over	over	IN
zkie-zero-2020	32	23	35	35	CD
zkie-zero-2020	32	24	%	%	NN
zkie-zero-2020	32	25	.	.	.
zkie-zero-2020	33	1	2	2	CD
zkie-zero-2020	33	2	Related	related	JJ
zkie-zero-2020	33	3	work	work	NN
zkie-zero-2020	33	4	In	in	IN
zkie-zero-2020	33	5	the	the	DT
zkie-zero-2020	33	6	following	following	NN
zkie-zero-2020	33	7	,	,	,
zkie-zero-2020	33	8	we	-PRON-	PRP
zkie-zero-2020	33	9	give	give	VBP
zkie-zero-2020	33	10	a	a	DT
zkie-zero-2020	33	11	broad	broad	JJ
zkie-zero-2020	33	12	overview	overview	NN
zkie-zero-2020	33	13	of	of	IN
zkie-zero-2020	33	14	exist-	exist-	NN
zkie-zero-2020	33	15	ing	ing	NNP
zkie-zero-2020	33	16	EL	EL	NNP
zkie-zero-2020	33	17	approaches	approach	NNS
zkie-zero-2020	33	18	,	,	,
zkie-zero-2020	33	19	annotation	annotation	NN
zkie-zero-2020	33	20	support	support	NN
zkie-zero-2020	33	21	and	and	CC
zkie-zero-2020	33	22	Human-	Human-	NNP
zkie-zero-2020	33	23	In	in	IN
zkie-zero-2020	33	24	-	-	HYPH
zkie-zero-2020	33	25	The	the	DT
zkie-zero-2020	33	26	-	-	HYPH
zkie-zero-2020	33	27	Loop	Loop	NNP
zkie-zero-2020	33	28	annotation	annotation	NN
zkie-zero-2020	33	29	.	.	.
zkie-zero-2020	34	1	Entity	Entity	NNP
zkie-zero-2020	34	2	Linking	Linking	NNP
zkie-zero-2020	34	3	describes	describe	VBZ
zkie-zero-2020	34	4	the	the	DT
zkie-zero-2020	34	5	task	task	NN
zkie-zero-2020	34	6	of	of	IN
zkie-zero-2020	34	7	disam-	disam-	NNP
zkie-zero-2020	34	8	biguating	biguating	NN
zkie-zero-2020	34	9	mentions	mention	NNS
zkie-zero-2020	34	10	in	in	IN
zkie-zero-2020	34	11	a	a	DT
zkie-zero-2020	34	12	text	text	NN
zkie-zero-2020	34	13	against	against	IN
zkie-zero-2020	34	14	a	a	DT
zkie-zero-2020	34	15	knowl-	knowl-	JJ
zkie-zero-2020	34	16	edge	edge	NN
zkie-zero-2020	34	17	base	base	NN
zkie-zero-2020	34	18	.	.	.
zkie-zero-2020	35	1	It	-PRON-	PRP
zkie-zero-2020	35	2	is	be	VBZ
zkie-zero-2020	35	3	typically	typically	RB
zkie-zero-2020	35	4	approached	approach	VBN
zkie-zero-2020	35	5	in	in	IN
zkie-zero-2020	35	6	three	three	CD
zkie-zero-2020	35	7	steps	step	NNS
zkie-zero-2020	35	8	:	:	:
zkie-zero-2020	35	9	1	1	LS
zkie-zero-2020	35	10	)	)	-RRB-
zkie-zero-2020	35	11	mention	mention	VB
zkie-zero-2020	35	12	detection	detection	NN
zkie-zero-2020	35	13	,	,	,
zkie-zero-2020	35	14	2	2	CD
zkie-zero-2020	35	15	)	)	-RRB-
zkie-zero-2020	35	16	candidate	candidate	NN
zkie-zero-2020	35	17	gener-	gener-	NN
zkie-zero-2020	35	18	ation	ation	NN
zkie-zero-2020	35	19	and	and	CC
zkie-zero-2020	35	20	3	3	LS
zkie-zero-2020	35	21	)	)	-RRB-
zkie-zero-2020	35	22	candidate	candidate	NN
zkie-zero-2020	35	23	ranking	rank	VBG
zkie-zero-2020	35	24	(	(	-LRB-
zkie-zero-2020	35	25	Shen	Shen	NNP
zkie-zero-2020	35	26	et	et	NNP
zkie-zero-2020	35	27	al	al	NNP
zkie-zero-2020	35	28	.	.	NNP
zkie-zero-2020	35	29	,	,	,
zkie-zero-2020	35	30	2015	2015	CD
zkie-zero-2020	35	31	)	)	-RRB-
zkie-zero-2020	35	32	(	(	-LRB-
zkie-zero-2020	35	33	Fig	fig	NN
zkie-zero-2020	35	34	.	.	.
zkie-zero-2020	36	1	2	2	LS
zkie-zero-2020	36	2	)	)	-RRB-
zkie-zero-2020	36	3	.	.	.
zkie-zero-2020	37	1	Mention	mention	VB
zkie-zero-2020	37	2	detection	detection	NN
zkie-zero-2020	37	3	most	most	RBS
zkie-zero-2020	37	4	often	often	RB
zkie-zero-2020	37	5	relies	rely	VBZ
zkie-zero-2020	37	6	either	either	CC
zkie-zero-2020	37	7	on	on	IN
zkie-zero-2020	37	8	gazetteers	gazetteer	NNS
zkie-zero-2020	37	9	or	or	CC
zkie-zero-2020	37	10	pretrained	pretrained	JJ
zkie-zero-2020	37	11	named	name	VBN
zkie-zero-2020	37	12	entity	entity	NN
zkie-zero-2020	37	13	recogniz-	recogniz-	JJ
zkie-zero-2020	37	14	ers	er	NNS
zkie-zero-2020	37	15	.	.	.
zkie-zero-2020	38	1	Candidate	candidate	NN
zkie-zero-2020	38	2	generation	generation	NN
zkie-zero-2020	38	3	either	either	CC
zkie-zero-2020	38	4	uses	use	VBZ
zkie-zero-2020	38	5	precompiled	precompile	VBN
zkie-zero-2020	38	6	candidate	candidate	NN
zkie-zero-2020	38	7	lists	list	NNS
zkie-zero-2020	38	8	derived	derive	VBN
zkie-zero-2020	38	9	from	from	IN
zkie-zero-2020	38	10	labeled	label	VBN
zkie-zero-2020	38	11	data	datum	NNS
zkie-zero-2020	38	12	or	or	CC
zkie-zero-2020	38	13	uses	use	VBZ
zkie-zero-2020	38	14	full	full	JJ
zkie-zero-2020	38	15	-	-	HYPH
zkie-zero-2020	38	16	text	text	NN
zkie-zero-2020	38	17	search	search	NN
zkie-zero-2020	38	18	.	.	.
zkie-zero-2020	39	1	Candidate	candidate	NN
zkie-zero-2020	39	2	ranking	rank	VBG
zkie-zero-2020	39	3	assigns	assign	NNS
zkie-zero-2020	39	4	each	each	DT
zkie-zero-2020	39	5	candidate	candidate	NN
zkie-zero-2020	39	6	a	a	DT
zkie-zero-2020	39	7	score	score	NN
zkie-zero-2020	39	8	,	,	,
zkie-zero-2020	39	9	then	then	RB
zkie-zero-2020	39	10	the	the	DT
zkie-zero-2020	39	11	candidate	candidate	NN
zkie-zero-2020	39	12	with	with	IN
zkie-zero-2020	39	13	the	the	DT
zkie-zero-2020	39	14	high-	high-	NN
zkie-zero-2020	39	15	est	est	NNP
zkie-zero-2020	39	16	score	score	NN
zkie-zero-2020	39	17	is	be	VBZ
zkie-zero-2020	39	18	returned	return	VBN
zkie-zero-2020	39	19	as	as	IN
zkie-zero-2020	39	20	the	the	DT
zkie-zero-2020	39	21	final	final	JJ
zkie-zero-2020	39	22	prediction	prediction	NN
zkie-zero-2020	39	23	.	.	.
zkie-zero-2020	40	1	Existing	exist	VBG
zkie-zero-2020	40	2	systems	system	NNS
zkie-zero-2020	40	3	rely	rely	VBP
zkie-zero-2020	40	4	on	on	IN
zkie-zero-2020	40	5	the	the	DT
zkie-zero-2020	40	6	availability	availability	NN
zkie-zero-2020	40	7	of	of	IN
zkie-zero-2020	40	8	certain	certain	JJ
zkie-zero-2020	40	9	resources	resource	NNS
zkie-zero-2020	40	10	like	like	IN
zkie-zero-2020	40	11	a	a	DT
zkie-zero-2020	40	12	large	large	JJ
zkie-zero-2020	40	13	Wikipedia	Wikipedia	NNP
zkie-zero-2020	40	14	as	as	RB
zkie-zero-2020	40	15	well	well	RB
zkie-zero-2020	40	16	as	as	IN
zkie-zero-2020	40	17	software	software	NN
zkie-zero-2020	40	18	tools	tool	NNS
zkie-zero-2020	40	19	and	and	CC
zkie-zero-2020	40	20	often	often	RB
zkie-zero-2020	40	21	are	be	VBP
zkie-zero-2020	40	22	restricted	restrict	VBN
zkie-zero-2020	40	23	in	in	IN
zkie-zero-2020	40	24	the	the	DT
zkie-zero-2020	40	25	knowledge	knowledge	NN
zkie-zero-2020	40	26	base	base	NN
zkie-zero-2020	40	27	they	-PRON-	PRP
zkie-zero-2020	40	28	can	can	MD
zkie-zero-2020	40	29	link	link	VB
zkie-zero-2020	40	30	to	to	IN
zkie-zero-2020	40	31	.	.	.
zkie-zero-2020	41	1	Off	off	IN
zkie-zero-2020	41	2	-	-	HYPH
zkie-zero-2020	41	3	the	the	DT
zkie-zero-2020	41	4	-	-	HYPH
zkie-zero-2020	41	5	shelf	shelf	NN
zkie-zero-2020	41	6	systems	system	NNS
zkie-zero-2020	41	7	like	like	IN
zkie-zero-2020	41	8	Dexter	Dexter	NNP
zkie-zero-2020	41	9	(	(	-LRB-
zkie-zero-2020	41	10	Ceccarelli	Ceccarelli	NNP
zkie-zero-2020	41	11	et	et	NNP
zkie-zero-2020	41	12	al	al	NNP
zkie-zero-2020	41	13	.	.	NNP
zkie-zero-2020	41	14	,	,	,
zkie-zero-2020	41	15	2013	2013	CD
zkie-zero-2020	41	16	)	)	-RRB-
zkie-zero-2020	41	17	,	,	,
zkie-zero-2020	41	18	DBPedia	DBPedia	NNP
zkie-zero-2020	41	19	Spotlight	Spotlight	NNP
zkie-zero-2020	41	20	(	(	-LRB-
zkie-zero-2020	41	21	Daiber	Daiber	NNP
zkie-zero-2020	41	22	et	et	FW
zkie-zero-2020	41	23	al	al	NNP
zkie-zero-2020	41	24	.	.	NNP
zkie-zero-2020	41	25	,	,	,
zkie-zero-2020	41	26	2013	2013	CD
zkie-zero-2020	41	27	)	)	-RRB-
zkie-zero-2020	41	28	and	and	CC
zkie-zero-2020	41	29	TagMe	TagMe	NNP
zkie-zero-2020	41	30	(	(	-LRB-
zkie-zero-2020	41	31	Ferragina	Ferragina	NNP
zkie-zero-2020	41	32	and	and	CC
zkie-zero-2020	41	33	Scaiella	Scaiella	NNP
zkie-zero-2020	41	34	,	,	,
zkie-zero-2020	41	35	2010	2010	CD
zkie-zero-2020	41	36	)	)	-RRB-
zkie-zero-2020	41	37	most	most	RBS
zkie-zero-2020	41	38	often	often	RB
zkie-zero-2020	41	39	can	can	MD
zkie-zero-2020	41	40	only	only	RB
zkie-zero-2020	41	41	link	link	VB
zkie-zero-2020	41	42	against	against	IN
zkie-zero-2020	41	43	Wikipedia	Wikipedia	NNP
zkie-zero-2020	41	44	or	or	CC
zkie-zero-2020	41	45	a	a	DT
zkie-zero-2020	41	46	related	related	JJ
zkie-zero-2020	41	47	knowledge	knowledge	NN
zkie-zero-2020	41	48	base	base	NN
zkie-zero-2020	41	49	like	like	IN
zkie-zero-2020	41	50	Wiki-	Wiki-	NNP
zkie-zero-2020	41	51	data	datum	NNS
zkie-zero-2020	41	52	or	or	CC
zkie-zero-2020	41	53	DBPedia	dbpedia	NN
zkie-zero-2020	41	54	.	.	.
zkie-zero-2020	42	1	They	-PRON-	PRP
zkie-zero-2020	42	2	require	require	VBP
zkie-zero-2020	42	3	good	good	JJ
zkie-zero-2020	42	4	Wikipedia	wikipedia	JJ
zkie-zero-2020	42	5	coverage	coverage	NN
zkie-zero-2020	42	6	for	for	IN
zkie-zero-2020	42	7	computing	compute	VBG
zkie-zero-2020	42	8	frequency	frequency	NN
zkie-zero-2020	42	9	statistics	statistic	NNS
zkie-zero-2020	42	10	like	like	IN
zkie-zero-2020	42	11	popularity	popularity	NN
zkie-zero-2020	42	12	,	,	,
zkie-zero-2020	42	13	view	view	NN
zkie-zero-2020	42	14	count	count	NN
zkie-zero-2020	42	15	or	or	CC
zkie-zero-2020	42	16	PageRank	PageRank	NNP
zkie-zero-2020	42	17	(	(	-LRB-
zkie-zero-2020	42	18	Guo	Guo	NNP
zkie-zero-2020	42	19	et	et	FW
zkie-zero-2020	42	20	al	al	NNP
zkie-zero-2020	42	21	.	.	NNP
zkie-zero-2020	42	22	,	,	,
zkie-zero-2020	42	23	2013	2013	CD
zkie-zero-2020	42	24	)	)	-RRB-
zkie-zero-2020	42	25	.	.	.
zkie-zero-2020	43	1	These	these	DT
zkie-zero-2020	43	2	features	feature	NNS
zkie-zero-2020	43	3	work	work	VBP
zkie-zero-2020	43	4	very	very	RB
zkie-zero-2020	43	5	well	well	RB
zkie-zero-2020	43	6	for	for	IN
zkie-zero-2020	43	7	stan-	stan-	NNP
zkie-zero-2020	43	8	dard	dard	NNP
zkie-zero-2020	43	9	datasets	dataset	NNS
zkie-zero-2020	43	10	due	due	IN
zkie-zero-2020	43	11	to	to	IN
zkie-zero-2020	43	12	their	-PRON-	PRP$
zkie-zero-2020	43	13	Zipfian	zipfian	JJ
zkie-zero-2020	43	14	distribution	distribution	NN
zkie-zero-2020	43	15	of	of	IN
zkie-zero-2020	43	16	entities	entity	NNS
zkie-zero-2020	43	17	,	,	,
zkie-zero-2020	43	18	leading	lead	VBG
zkie-zero-2020	43	19	to	to	IN
zkie-zero-2020	43	20	high	high	JJ
zkie-zero-2020	43	21	reported	report	VBN
zkie-zero-2020	43	22	scores	score	NNS
zkie-zero-2020	43	23	on	on	IN
zkie-zero-2020	43	24	state-	state-	NNP
zkie-zero-2020	43	25	of	of	IN
zkie-zero-2020	43	26	-	-	:
zkie-zero-2020	43	27	the	the	DT
zkie-zero-2020	43	28	art	art	NN
zkie-zero-2020	43	29	datasets	dataset	NNS
zkie-zero-2020	43	30	(	(	-LRB-
zkie-zero-2020	43	31	Ilievski	Ilievski	NNP
zkie-zero-2020	43	32	et	et	NNP
zkie-zero-2020	43	33	al	al	NNP
zkie-zero-2020	43	34	.	.	NNP
zkie-zero-2020	43	35	,	,	,
zkie-zero-2020	43	36	2018	2018	CD
zkie-zero-2020	43	37	;	;	:
zkie-zero-2020	43	38	Milne	Milne	NNP
zkie-zero-2020	43	39	and	and	CC
zkie-zero-2020	43	40	Witten	Witten	NNP
zkie-zero-2020	43	41	,	,	,
zkie-zero-2020	43	42	2008	2008	CD
zkie-zero-2020	43	43	)	)	-RRB-
zkie-zero-2020	43	44	.	.	.
zkie-zero-2020	44	1	However	however	RB
zkie-zero-2020	44	2	,	,	,
zkie-zero-2020	44	3	these	these	DT
zkie-zero-2020	44	4	systems	system	NNS
zkie-zero-2020	44	5	are	be	VBP
zkie-zero-2020	44	6	rarely	rarely	RB
zkie-zero-2020	44	7	applied	apply	VBN
zkie-zero-2020	44	8	out	out	JJ
zkie-zero-2020	44	9	-	-	HYPH
zkie-zero-2020	44	10	of	of	IN
zkie-zero-2020	44	11	-	-	HYPH
zkie-zero-2020	44	12	domain	domain	NN
zkie-zero-2020	44	13	such	such	JJ
zkie-zero-2020	44	14	as	as	IN
zkie-zero-2020	44	15	in	in	IN
zkie-zero-2020	44	16	digital	digital	JJ
zkie-zero-2020	44	17	humanities	humanity	NNS
zkie-zero-2020	44	18	or	or	CC
zkie-zero-2020	44	19	classical	classical	JJ
zkie-zero-2020	44	20	studies	study	NNS
zkie-zero-2020	44	21	.	.	.
zkie-zero-2020	45	1	Compared	compare	VBN
zkie-zero-2020	45	2	to	to	IN
zkie-zero-2020	45	3	state	state	NN
zkie-zero-2020	45	4	-	-	HYPH
zkie-zero-2020	45	5	of	of	IN
zkie-zero-2020	45	6	-	-	HYPH
zkie-zero-2020	45	7	the	the	DT
zkie-zero-2020	45	8	-	-	HYPH
zkie-zero-2020	45	9	art	art	NN
zkie-zero-2020	45	10	approaches	approach	NNS
zkie-zero-2020	45	11	,	,	,
zkie-zero-2020	45	12	only	only	RB
zkie-zero-2020	45	13	a	a	DT
zkie-zero-2020	45	14	limited	limited	JJ
zkie-zero-2020	45	15	amount	amount	NN
zkie-zero-2020	45	16	of	of	IN
zkie-zero-2020	45	17	research	research	NN
zkie-zero-2020	45	18	has	have	VBZ
zkie-zero-2020	45	19	been	be	VBN
zkie-zero-2020	45	20	performed	perform	VBN
zkie-zero-2020	45	21	on	on	IN
zkie-zero-2020	45	22	entity	entity	NN
zkie-zero-2020	45	23	linking	link	VBG
zkie-zero-2020	45	24	against	against	IN
zkie-zero-2020	45	25	domain-	domain-	JJ
zkie-zero-2020	45	26	specific	specific	JJ
zkie-zero-2020	45	27	knowledge	knowledge	NN
zkie-zero-2020	45	28	bases	basis	NNS
zkie-zero-2020	45	29	.	.	.
zkie-zero-2020	46	1	AGDISTIS	AGDISTIS	NNP
zkie-zero-2020	46	2	(	(	-LRB-
zkie-zero-2020	46	3	Usbeck	Usbeck	NNP
zkie-zero-2020	46	4	et	et	NNP
zkie-zero-2020	46	5	al	al	NNP
zkie-zero-2020	46	6	.	.	NNP
zkie-zero-2020	46	7	,	,	,
zkie-zero-2020	46	8	2014	2014	CD
zkie-zero-2020	46	9	)	)	-RRB-
zkie-zero-2020	46	10	developed	develop	VBD
zkie-zero-2020	46	11	a	a	DT
zkie-zero-2020	46	12	knowledge	knowledge	NN
zkie-zero-2020	46	13	-	-	HYPH
zkie-zero-2020	46	14	base	base	NN
zkie-zero-2020	46	15	-	-	HYPH
zkie-zero-2020	46	16	agnostic	agnostic	JJ
zkie-zero-2020	46	17	approach	approach	NN
zkie-zero-2020	46	18	based	base	VBN
zkie-zero-2020	46	19	on	on	IN
zkie-zero-2020	46	20	the	the	DT
zkie-zero-2020	46	21	HITS	HITS	NNP
zkie-zero-2020	46	22	algorithm	algorithm	NN
zkie-zero-2020	46	23	.	.	.
zkie-zero-2020	47	1	The	the	DT
zkie-zero-2020	47	2	men-	men-	NN
zkie-zero-2020	47	3	tion	tion	NN
zkie-zero-2020	47	4	detection	detection	NN
zkie-zero-2020	47	5	relies	rely	VBZ
zkie-zero-2020	47	6	on	on	IN
zkie-zero-2020	47	7	gazetteers	gazetteer	NNS
zkie-zero-2020	47	8	compiled	compile	VBN
zkie-zero-2020	47	9	from	from	IN
zkie-zero-2020	47	10	re-	re-	JJ
zkie-zero-2020	47	11	sources	source	NNS
zkie-zero-2020	47	12	like	like	IN
zkie-zero-2020	47	13	Wikipedia	Wikipedia	NNP
zkie-zero-2020	47	14	and	and	CC
zkie-zero-2020	47	15	thereby	thereby	RB
zkie-zero-2020	47	16	performs	perform	VBZ
zkie-zero-2020	47	17	string	string	NN
zkie-zero-2020	47	18	matching	matching	NN
zkie-zero-2020	47	19	.	.	.
zkie-zero-2020	48	1	Brando	Brando	NNP
zkie-zero-2020	48	2	et	et	FW
zkie-zero-2020	48	3	al	al	NNP
zkie-zero-2020	48	4	.	.	.
zkie-zero-2020	49	1	(	(	-LRB-
zkie-zero-2020	49	2	2016	2016	CD
zkie-zero-2020	49	3	)	)	-RRB-
zkie-zero-2020	49	4	propose	propose	NN
zkie-zero-2020	49	5	REDEN	REDEN	NNP
zkie-zero-2020	49	6	,	,	,
zkie-zero-2020	49	7	an	an	DT
zkie-zero-2020	49	8	approach	approach	NN
zkie-zero-2020	49	9	based	base	VBN
zkie-zero-2020	49	10	on	on	IN
zkie-zero-2020	49	11	graph	graph	NN
zkie-zero-2020	49	12	centrality	centrality	NN
zkie-zero-2020	49	13	to	to	TO
zkie-zero-2020	49	14	link	link	VB
zkie-zero-2020	49	15	French	french	JJ
zkie-zero-2020	49	16	authors	author	NNS
zkie-zero-2020	49	17	to	to	IN
zkie-zero-2020	49	18	literary	literary	JJ
zkie-zero-2020	49	19	criticism	criticism	NN
zkie-zero-2020	49	20	texts	text	NNS
zkie-zero-2020	49	21	.	.	.
zkie-zero-2020	50	1	It	-PRON-	PRP
zkie-zero-2020	50	2	requires	require	VBZ
zkie-zero-2020	50	3	addi-	addi-	JJ
zkie-zero-2020	50	4	tional	tional	JJ
zkie-zero-2020	50	5	linked	link	VBN
zkie-zero-2020	50	6	data	datum	NNS
zkie-zero-2020	50	7	that	that	WDT
zkie-zero-2020	50	8	is	be	VBZ
zkie-zero-2020	50	9	aligned	align	VBN
zkie-zero-2020	50	10	with	with	IN
zkie-zero-2020	50	11	the	the	DT
zkie-zero-2020	50	12	custom	custom	NN
zkie-zero-2020	50	13	knowledge	knowledge	NN
zkie-zero-2020	50	14	base	base	NN
zkie-zero-2020	50	15	–	–	:
zkie-zero-2020	50	16	they	-PRON-	PRP
zkie-zero-2020	50	17	use	use	VBP
zkie-zero-2020	50	18	DBPedia	DBPedia	NNP
zkie-zero-2020	50	19	.	.	.
zkie-zero-2020	51	1	As	as	IN
zkie-zero-2020	51	2	we	-PRON-	PRP
zkie-zero-2020	51	3	work	work	VBP
zkie-zero-2020	51	4	in	in	IN
zkie-zero-2020	51	5	a	a	DT
zkie-zero-2020	51	6	domain	domain	NN
zkie-zero-2020	51	7	-	-	HYPH
zkie-zero-2020	51	8	specific	specific	JJ
zkie-zero-2020	51	9	low	low	JJ
zkie-zero-2020	51	10	resource	resource	NN
zkie-zero-2020	51	11	setting	setting	NN
zkie-zero-2020	51	12	,	,	,
zkie-zero-2020	51	13	access	access	NN
zkie-zero-2020	51	14	to	to	IN
zkie-zero-2020	51	15	large	large	JJ
zkie-zero-2020	51	16	corpora	corpora	NN
zkie-zero-2020	51	17	which	which	WDT
zkie-zero-2020	51	18	can	can	MD
zkie-zero-2020	51	19	be	be	VB
zkie-zero-2020	51	20	used	use	VBN
zkie-zero-2020	51	21	to	to	TO
zkie-zero-2020	51	22	compute	compute	VB
zkie-zero-2020	51	23	pop-	pop-	NN
zkie-zero-2020	51	24	ularity	ularity	NN
zkie-zero-2020	51	25	priors	prior	NNS
zkie-zero-2020	51	26	is	be	VBZ
zkie-zero-2020	51	27	limited	limited	JJ
zkie-zero-2020	51	28	.	.	.
zkie-zero-2020	52	1	We	-PRON-	PRP
zkie-zero-2020	52	2	do	do	VBP
zkie-zero-2020	52	3	not	not	RB
zkie-zero-2020	52	4	have	have	VB
zkie-zero-2020	52	5	suitable	suitable	JJ
zkie-zero-2020	52	6	named	name	VBN
zkie-zero-2020	52	7	entity	entity	NN
zkie-zero-2020	52	8	linking	linking	NN
zkie-zero-2020	52	9	tools	tool	NNS
zkie-zero-2020	52	10	,	,	,
zkie-zero-2020	52	11	gazetteers	gazetteer	NNS
zkie-zero-2020	52	12	or	or	CC
zkie-zero-2020	52	13	a	a	DT
zkie-zero-2020	52	14	sufficient	sufficient	JJ
zkie-zero-2020	52	15	amount	amount	NN
zkie-zero-2020	52	16	of	of	IN
zkie-zero-2020	52	17	labeled	label	VBN
zkie-zero-2020	52	18	training	training	NN
zkie-zero-2020	52	19	data	datum	NNS
zkie-zero-2020	52	20	.	.	.
zkie-zero-2020	53	1	Therefore	therefore	RB
zkie-zero-2020	53	2	,	,	,
zkie-zero-2020	53	3	it	-PRON-	PRP
zkie-zero-2020	53	4	is	be	VBZ
zkie-zero-2020	53	5	challenging	challenge	VBG
zkie-zero-2020	53	6	to	to	TO
zkie-zero-2020	53	7	use	use	VB
zkie-zero-2020	53	8	state	state	NN
zkie-zero-2020	53	9	of	of	IN
zkie-zero-2020	53	10	the	the	DT
zkie-zero-2020	53	11	art	art	NN
zkie-zero-2020	53	12	systems	system	NNS
zkie-zero-2020	53	13	.	.	.
zkie-zero-2020	54	1	Human	human	JJ
zkie-zero-2020	54	2	-	-	HYPH
zkie-zero-2020	54	3	in	in	IN
zkie-zero-2020	54	4	-	-	HYPH
zkie-zero-2020	54	5	the	the	DT
zkie-zero-2020	54	6	-	-	HYPH
zkie-zero-2020	54	7	loop	loop	NN
zkie-zero-2020	54	8	annotation	annotation	NN
zkie-zero-2020	54	9	HITL	HITL	NNP
zkie-zero-2020	54	10	machine	machine	NN
zkie-zero-2020	54	11	learning	learning	NN
zkie-zero-2020	54	12	describes	describe	VBZ
zkie-zero-2020	54	13	an	an	DT
zkie-zero-2020	54	14	interactive	interactive	JJ
zkie-zero-2020	54	15	scenario	scenario	NN
zkie-zero-2020	54	16	where	where	WRB
zkie-zero-2020	54	17	a	a	DT
zkie-zero-2020	54	18	machine	machine	NN
zkie-zero-2020	54	19	learning	learn	VBG
zkie-zero-2020	54	20	(	(	-LRB-
zkie-zero-2020	54	21	ML	ML	NNP
zkie-zero-2020	54	22	)	)	-RRB-
zkie-zero-2020	54	23	system	system	NN
zkie-zero-2020	54	24	and	and	CC
zkie-zero-2020	54	25	a	a	DT
zkie-zero-2020	54	26	human	human	JJ
zkie-zero-2020	54	27	work	work	NN
zkie-zero-2020	54	28	together	together	RB
zkie-zero-2020	54	29	to	to	TO
zkie-zero-2020	54	30	improve	improve	VB
zkie-zero-2020	54	31	their	-PRON-	PRP$
zkie-zero-2020	54	32	performance	performance	NN
zkie-zero-2020	54	33	.	.	.
zkie-zero-2020	55	1	The	the	DT
zkie-zero-2020	55	2	ML	ML	NNP
zkie-zero-2020	55	3	system	system	NN
zkie-zero-2020	55	4	gives	give	VBZ
zkie-zero-2020	55	5	predictions	prediction	NNS
zkie-zero-2020	55	6	,	,	,
zkie-zero-2020	55	7	and	and	CC
zkie-zero-2020	55	8	the	the	DT
zkie-zero-2020	55	9	human	human	JJ
zkie-zero-2020	55	10	corrects	correct	VBZ
zkie-zero-2020	55	11	if	if	IN
zkie-zero-2020	55	12	they	-PRON-	PRP
zkie-zero-2020	55	13	are	be	VBP
zkie-zero-2020	55	14	wrong	wrong	JJ
zkie-zero-2020	55	15	and	and	CC
zkie-zero-2020	55	16	helps	help	VBZ
zkie-zero-2020	55	17	to	to	TO
zkie-zero-2020	55	18	spot	spot	VB
zkie-zero-2020	55	19	things	thing	NNS
zkie-zero-2020	55	20	that	that	WDT
zkie-zero-2020	55	21	have	have	VBP
zkie-zero-2020	55	22	been	be	VBN
zkie-zero-2020	55	23	overlooked	overlook	VBN
zkie-zero-2020	55	24	by	by	IN
zkie-zero-2020	55	25	the	the	DT
zkie-zero-2020	55	26	machine	machine	NN
zkie-zero-2020	55	27	.	.	.
zkie-zero-2020	56	1	The	the	DT
zkie-zero-2020	56	2	sys-	sys-	NN
zkie-zero-2020	56	3	tem	tem	NN
zkie-zero-2020	56	4	uses	use	VBZ
zkie-zero-2020	56	5	this	this	DT
zkie-zero-2020	56	6	feedback	feedback	NN
zkie-zero-2020	56	7	to	to	TO
zkie-zero-2020	56	8	improve	improve	VB
zkie-zero-2020	56	9	,	,	,
zkie-zero-2020	56	10	leading	lead	VBG
zkie-zero-2020	56	11	to	to	IN
zkie-zero-2020	56	12	bet-	bet-	JJ
zkie-zero-2020	56	13	ter	ter	NN
zkie-zero-2020	56	14	predictions	prediction	NNS
zkie-zero-2020	56	15	and	and	CC
zkie-zero-2020	56	16	thereby	thereby	RB
zkie-zero-2020	56	17	reducing	reduce	VBG
zkie-zero-2020	56	18	the	the	DT
zkie-zero-2020	56	19	effort	effort	NN
zkie-zero-2020	56	20	of	of	IN
zkie-zero-2020	56	21	the	the	DT
zkie-zero-2020	56	22	human	human	NN
zkie-zero-2020	56	23	.	.	.
zkie-zero-2020	57	1	In	in	IN
zkie-zero-2020	57	2	natural	natural	JJ
zkie-zero-2020	57	3	language	language	NN
zkie-zero-2020	57	4	processing	processing	NN
zkie-zero-2020	57	5	,	,	,
zkie-zero-2020	57	6	it	-PRON-	PRP
zkie-zero-2020	57	7	has	have	VBZ
zkie-zero-2020	57	8	been	be	VBN
zkie-zero-2020	57	9	applied	apply	VBN
zkie-zero-2020	57	10	in	in	IN
zkie-zero-2020	57	11	scenarios	scenario	NNS
zkie-zero-2020	57	12	like	like	IN
zkie-zero-2020	57	13	interactive	interactive	NNP
zkie-zero-2020	57	14	text	text	NNP
zkie-zero-2020	57	15	sum-	sum-	NNP
zkie-zero-2020	57	16	marization	marization	NN
zkie-zero-2020	57	17	(	(	-LRB-
zkie-zero-2020	57	18	Gao	Gao	NNP
zkie-zero-2020	57	19	et	et	NNP
zkie-zero-2020	57	20	al	al	NNP
zkie-zero-2020	57	21	.	.	NNP
zkie-zero-2020	57	22	,	,	,
zkie-zero-2020	57	23	2018	2018	CD
zkie-zero-2020	57	24	)	)	-RRB-
zkie-zero-2020	57	25	,	,	,
zkie-zero-2020	57	26	parsing	parse	VBG
zkie-zero-2020	57	27	(	(	-LRB-
zkie-zero-2020	57	28	He	-PRON-	PRP
zkie-zero-2020	57	29	et	et	FW
zkie-zero-2020	57	30	al	al	NNP
zkie-zero-2020	57	31	.	.	NNP
zkie-zero-2020	57	32	,	,	,
zkie-zero-2020	57	33	2016	2016	CD
zkie-zero-2020	57	34	)	)	-RRB-
zkie-zero-2020	57	35	or	or	CC
zkie-zero-2020	57	36	data	datum	NNS
zkie-zero-2020	57	37	generation	generation	NN
zkie-zero-2020	57	38	(	(	-LRB-
zkie-zero-2020	57	39	Wallace	Wallace	NNP
zkie-zero-2020	57	40	et	et	FW
zkie-zero-2020	57	41	al	al	NNP
zkie-zero-2020	57	42	.	.	NNP
zkie-zero-2020	57	43	,	,	,
zkie-zero-2020	57	44	2019	2019	CD
zkie-zero-2020	57	45	)	)	-RRB-
zkie-zero-2020	57	46	.	.	.
zkie-zero-2020	58	1	Regarding	regard	VBG
zkie-zero-2020	58	2	machine	machine	NN
zkie-zero-2020	58	3	-	-	HYPH
zkie-zero-2020	58	4	learning	learn	VBG
zkie-zero-2020	58	5	assisted	assisted	JJ
zkie-zero-2020	58	6	annotation	annotation	NN
zkie-zero-2020	58	7	,	,	,
zkie-zero-2020	58	8	Yimam	Yimam	NNP
zkie-zero-2020	58	9	et	et	NNP
zkie-zero-2020	58	10	al	al	NNP
zkie-zero-2020	58	11	.	.	.
zkie-zero-2020	59	1	(	(	-LRB-
zkie-zero-2020	59	2	2014	2014	CD
zkie-zero-2020	59	3	)	)	-RRB-
zkie-zero-2020	59	4	propose	propose	VB
zkie-zero-2020	59	5	an	an	DT
zkie-zero-2020	59	6	annotation	annotation	NN
zkie-zero-2020	59	7	editor	editor	NN
zkie-zero-2020	59	8	that	that	WDT
zkie-zero-2020	59	9	during	during	IN
zkie-zero-2020	59	10	annotation	annotation	NN
zkie-zero-2020	59	11	,	,	,
zkie-zero-2020	59	12	interactively	interactively	RB
zkie-zero-2020	59	13	trains	train	VBZ
zkie-zero-2020	59	14	a	a	DT
zkie-zero-2020	59	15	model	model	NN
zkie-zero-2020	59	16	using	use	VBG
zkie-zero-2020	59	17	annotations	annotation	NNS
zkie-zero-2020	59	18	made	make	VBN
zkie-zero-2020	59	19	by	by	IN
zkie-zero-2020	59	20	the	the	DT
zkie-zero-2020	59	21	user	user	NN
zkie-zero-2020	59	22	.	.	.
zkie-zero-2020	60	1	They	-PRON-	PRP
zkie-zero-2020	60	2	use	use	VBP
zkie-zero-2020	60	3	string	string	NN
zkie-zero-2020	60	4	matching	matching	NN
zkie-zero-2020	60	5	and	and	CC
zkie-zero-2020	60	6	MIRA	MIRA	NNP
zkie-zero-2020	60	7	(	(	-LRB-
zkie-zero-2020	60	8	Crammer	Crammer	NNP
zkie-zero-2020	60	9	and	and	CC
zkie-zero-2020	60	10	Singer	Singer	NNP
zkie-zero-2020	60	11	,	,	,
zkie-zero-2020	60	12	2003	2003	CD
zkie-zero-2020	60	13	)	)	-RRB-
zkie-zero-2020	60	14	as	as	IN
zkie-zero-2020	60	15	recommenders	recommender	NNS
zkie-zero-2020	60	16	,	,	,
zkie-zero-2020	60	17	evaluate	evaluate	VB
zkie-zero-2020	60	18	on	on	IN
zkie-zero-2020	60	19	POS	POS	NNP
zkie-zero-2020	60	20	and	and	CC
zkie-zero-2020	60	21	NER	NER	NNP
zkie-zero-2020	60	22	anno-	anno-	JJ
zkie-zero-2020	60	23	tation	tation	NN
zkie-zero-2020	60	24	and	and	CC
zkie-zero-2020	60	25	show	show	NN
zkie-zero-2020	60	26	improvement	improvement	NN
zkie-zero-2020	60	27	in	in	IN
zkie-zero-2020	60	28	annotation	annotation	NN
zkie-zero-2020	60	29	speed	speed	NN
zkie-zero-2020	60	30	.	.	.
zkie-zero-2020	61	1	TASTY	TASTY	NNP
zkie-zero-2020	61	2	(	(	-LRB-
zkie-zero-2020	61	3	Arnold	Arnold	NNP
zkie-zero-2020	61	4	et	et	NNP
zkie-zero-2020	61	5	al	al	NNP
zkie-zero-2020	61	6	.	.	NNP
zkie-zero-2020	61	7	,	,	,
zkie-zero-2020	61	8	2016	2016	CD
zkie-zero-2020	61	9	)	)	-RRB-
zkie-zero-2020	61	10	is	be	VBZ
zkie-zero-2020	61	11	a	a	DT
zkie-zero-2020	61	12	system	system	NN
zkie-zero-2020	61	13	that	that	WDT
zkie-zero-2020	61	14	is	be	VBZ
zkie-zero-2020	61	15	able	able	JJ
zkie-zero-2020	61	16	to	to	TO
zkie-zero-2020	61	17	perform	perform	VB
zkie-zero-2020	61	18	EL	el	NN
zkie-zero-2020	61	19	against	against	IN
zkie-zero-2020	61	20	Wikipedia	Wikipedia	NNP
zkie-zero-2020	61	21	on	on	IN
zkie-zero-2020	61	22	the	the	DT
zkie-zero-2020	61	23	fly	fly	NN
zkie-zero-2020	61	24	while	while	IN
zkie-zero-2020	61	25	typing	type	VBG
zkie-zero-2020	61	26	a	a	DT
zkie-zero-2020	61	27	document	document	NN
zkie-zero-2020	61	28	.	.	.
zkie-zero-2020	62	1	A	a	DT
zkie-zero-2020	62	2	pretrained	pretrained	JJ
zkie-zero-2020	62	3	neural	neural	JJ
zkie-zero-2020	62	4	se-	se-	NN
zkie-zero-2020	62	5	quence	quence	NN
zkie-zero-2020	62	6	tagger	tagger	NNP
zkie-zero-2020	62	7	is	be	VBZ
zkie-zero-2020	62	8	being	be	VBG
zkie-zero-2020	62	9	used	use	VBN
zkie-zero-2020	62	10	that	that	IN
zkie-zero-2020	62	11	performs	perform	VBZ
zkie-zero-2020	62	12	mention	mention	VB
zkie-zero-2020	62	13	detection	detection	NN
zkie-zero-2020	62	14	.	.	.
zkie-zero-2020	63	1	Candidates	candidate	NNS
zkie-zero-2020	63	2	are	be	VBP
zkie-zero-2020	63	3	precomputed	precompute	VBN
zkie-zero-2020	63	4	and	and	CC
zkie-zero-2020	63	5	the	the	DT
zkie-zero-2020	63	6	candidate	candidate	NN
zkie-zero-2020	63	7	is	be	VBZ
zkie-zero-2020	63	8	chosen	choose	VBN
zkie-zero-2020	63	9	that	that	WDT
zkie-zero-2020	63	10	has	have	VBZ
zkie-zero-2020	63	11	the	the	DT
zkie-zero-2020	63	12	highest	high	JJS
zkie-zero-2020	63	13	text	text	NN
zkie-zero-2020	63	14	sim-	sim-	NNS
zkie-zero-2020	63	15	6984	6984	CD
zkie-zero-2020	63	16	Figure	figure	NN
zkie-zero-2020	63	17	2	2	CD
zkie-zero-2020	63	18	:	:	:
zkie-zero-2020	63	19	Entity	entity	NN
zkie-zero-2020	63	20	linking	linking	NN
zkie-zero-2020	63	21	pipeline	pipeline	NN
zkie-zero-2020	63	22	:	:	:
zkie-zero-2020	63	23	First	first	RB
zkie-zero-2020	63	24	,	,	,
zkie-zero-2020	63	25	mentions	mention	NNS
zkie-zero-2020	63	26	of	of	IN
zkie-zero-2020	63	27	entities	entity	NNS
zkie-zero-2020	63	28	in	in	IN
zkie-zero-2020	63	29	the	the	DT
zkie-zero-2020	63	30	text	text	NN
zkie-zero-2020	63	31	need	need	NN
zkie-zero-2020	63	32	to	to	TO
zkie-zero-2020	63	33	be	be	VB
zkie-zero-2020	63	34	found	find	VBN
zkie-zero-2020	63	35	.	.	.
zkie-zero-2020	64	1	Then	then	RB
zkie-zero-2020	64	2	,	,	,
zkie-zero-2020	64	3	given	give	VBN
zkie-zero-2020	64	4	a	a	DT
zkie-zero-2020	64	5	mention	mention	NN
zkie-zero-2020	64	6	,	,	,
zkie-zero-2020	64	7	candidate	candidate	NN
zkie-zero-2020	64	8	entities	entity	NNS
zkie-zero-2020	64	9	are	be	VBP
zkie-zero-2020	64	10	generated	generate	VBN
zkie-zero-2020	64	11	.	.	.
zkie-zero-2020	65	1	Finally	finally	RB
zkie-zero-2020	65	2	,	,	,
zkie-zero-2020	65	3	entities	entity	NNS
zkie-zero-2020	65	4	are	be	VBP
zkie-zero-2020	65	5	ranked	rank	VBN
zkie-zero-2020	65	6	and	and	CC
zkie-zero-2020	65	7	the	the	DT
zkie-zero-2020	65	8	top	top	JJ
zkie-zero-2020	65	9	entity	entity	NN
zkie-zero-2020	65	10	is	be	VBZ
zkie-zero-2020	65	11	chosen	choose	VBN
zkie-zero-2020	65	12	.	.	.
zkie-zero-2020	66	1	ilarity	ilarity	NN
zkie-zero-2020	66	2	.	.	.
zkie-zero-2020	67	1	The	the	DT
zkie-zero-2020	67	2	system	system	NN
zkie-zero-2020	67	3	updates	update	VBZ
zkie-zero-2020	67	4	its	-PRON-	PRP$
zkie-zero-2020	67	5	suggestions	suggestion	NNS
zkie-zero-2020	67	6	after	after	IN
zkie-zero-2020	67	7	interactions	interaction	NNS
zkie-zero-2020	67	8	such	such	JJ
zkie-zero-2020	67	9	as	as	IN
zkie-zero-2020	67	10	writing	writing	NN
zkie-zero-2020	67	11	,	,	,
zkie-zero-2020	67	12	rephrasing	rephrase	VBG
zkie-zero-2020	67	13	,	,	,
zkie-zero-2020	67	14	removing	remove	VBG
zkie-zero-2020	67	15	or	or	CC
zkie-zero-2020	67	16	correcting	correct	VBG
zkie-zero-2020	67	17	suggested	suggest	VBN
zkie-zero-2020	67	18	entity	entity	NN
zkie-zero-2020	67	19	links	link	NNS
zkie-zero-2020	67	20	.	.	.
zkie-zero-2020	68	1	Corrections	correction	NNS
zkie-zero-2020	68	2	are	be	VBP
zkie-zero-2020	68	3	used	use	VBN
zkie-zero-2020	68	4	as	as	IN
zkie-zero-2020	68	5	training	training	NN
zkie-zero-2020	68	6	data	datum	NNS
zkie-zero-2020	68	7	for	for	IN
zkie-zero-2020	68	8	the	the	DT
zkie-zero-2020	68	9	neural	neural	JJ
zkie-zero-2020	68	10	model	model	NN
zkie-zero-2020	68	11	.	.	.
zkie-zero-2020	69	1	How-	How-	NNP
zkie-zero-2020	69	2	ever	ever	RB
zkie-zero-2020	69	3	,	,	,
zkie-zero-2020	69	4	due	due	IN
zkie-zero-2020	69	5	to	to	IN
zkie-zero-2020	69	6	the	the	DT
zkie-zero-2020	69	7	following	follow	VBG
zkie-zero-2020	69	8	reasons	reason	NNS
zkie-zero-2020	69	9	,	,	,
zkie-zero-2020	69	10	it	-PRON-	PRP
zkie-zero-2020	69	11	is	be	VBZ
zkie-zero-2020	69	12	not	not	RB
zkie-zero-2020	69	13	yet	yet	RB
zkie-zero-2020	69	14	suit-	suit-	VBN
zkie-zero-2020	69	15	able	able	JJ
zkie-zero-2020	69	16	for	for	IN
zkie-zero-2020	69	17	our	-PRON-	PRP$
zkie-zero-2020	69	18	scenario	scenario	NN
zkie-zero-2020	69	19	.	.	.
zkie-zero-2020	70	1	In	in	IN
zkie-zero-2020	70	2	order	order	NN
zkie-zero-2020	70	3	to	to	TO
zkie-zero-2020	70	4	overcome	overcome	VB
zkie-zero-2020	70	5	the	the	DT
zkie-zero-2020	70	6	cold	cold	JJ
zkie-zero-2020	70	7	start	start	NN
zkie-zero-2020	70	8	problem	problem	NN
zkie-zero-2020	70	9	,	,	,
zkie-zero-2020	70	10	it	-PRON-	PRP
zkie-zero-2020	70	11	needs	need	VBZ
zkie-zero-2020	70	12	annotated	annotate	VBN
zkie-zero-2020	70	13	training	train	VBG
zkie-zero-2020	70	14	data	datum	NNS
zkie-zero-2020	70	15	in	in	IN
zkie-zero-2020	70	16	addition	addition	NN
zkie-zero-2020	70	17	to	to	IN
zkie-zero-2020	70	18	a	a	DT
zkie-zero-2020	70	19	precomputed	precomputed	JJ
zkie-zero-2020	70	20	index	index	NN
zkie-zero-2020	70	21	for	for	IN
zkie-zero-2020	70	22	candidate	candidate	NN
zkie-zero-2020	70	23	generation	generation	NN
zkie-zero-2020	70	24	.	.	.
zkie-zero-2020	71	1	It	-PRON-	PRP
zkie-zero-2020	71	2	also	also	RB
zkie-zero-2020	71	3	only	only	RB
zkie-zero-2020	71	4	links	link	VBZ
zkie-zero-2020	71	5	against	against	IN
zkie-zero-2020	71	6	Wikipedia	Wikipedia	NNP
zkie-zero-2020	71	7	.	.	.
zkie-zero-2020	72	1	3	3	CD
zkie-zero-2020	72	2	Architecture	architecture	NN
zkie-zero-2020	72	3	The	the	DT
zkie-zero-2020	72	4	following	follow	VBG
zkie-zero-2020	72	5	section	section	NN
zkie-zero-2020	72	6	describes	describe	VBZ
zkie-zero-2020	72	7	the	the	DT
zkie-zero-2020	72	8	three	three	CD
zkie-zero-2020	72	9	com-	com-	NN
zkie-zero-2020	72	10	ponents	ponent	NNS
zkie-zero-2020	72	11	of	of	IN
zkie-zero-2020	72	12	our	-PRON-	PRP$
zkie-zero-2020	72	13	annotation	annotation	NN
zkie-zero-2020	72	14	framework	framework	NN
zkie-zero-2020	72	15	,	,	,
zkie-zero-2020	72	16	following	follow	VBG
zkie-zero-2020	72	17	the	the	DT
zkie-zero-2020	72	18	standard	standard	JJ
zkie-zero-2020	72	19	entity	entity	NN
zkie-zero-2020	72	20	linking	linking	NN
zkie-zero-2020	72	21	pipeline	pipeline	NN
zkie-zero-2020	72	22	(	(	-LRB-
zkie-zero-2020	72	23	see	see	VB
zkie-zero-2020	72	24	Fig	Fig	NNP
zkie-zero-2020	72	25	.	.	.
zkie-zero-2020	73	1	2	2	LS
zkie-zero-2020	73	2	)	)	-RRB-
zkie-zero-2020	73	3	.	.	.
zkie-zero-2020	74	1	Throughout	throughout	IN
zkie-zero-2020	74	2	this	this	DT
zkie-zero-2020	74	3	work	work	NN
zkie-zero-2020	74	4	,	,	,
zkie-zero-2020	74	5	we	-PRON-	PRP
zkie-zero-2020	74	6	will	will	MD
zkie-zero-2020	74	7	mainly	mainly	RB
zkie-zero-2020	74	8	focus	focus	VB
zkie-zero-2020	74	9	on	on	IN
zkie-zero-2020	74	10	the	the	DT
zkie-zero-2020	74	11	candidate	candidate	NN
zkie-zero-2020	74	12	Ranking	rank	VBG
zkie-zero-2020	74	13	step	step	NN
zkie-zero-2020	74	14	.	.	.
zkie-zero-2020	75	1	We	-PRON-	PRP
zkie-zero-2020	75	2	call	call	VBP
zkie-zero-2020	75	3	the	the	DT
zkie-zero-2020	75	4	text	text	NN
zkie-zero-2020	75	5	span	span	NN
zkie-zero-2020	75	6	which	which	WDT
zkie-zero-2020	75	7	contains	contain	VBZ
zkie-zero-2020	75	8	an	an	DT
zkie-zero-2020	75	9	entity	entity	NN
zkie-zero-2020	75	10	the	the	DT
zkie-zero-2020	75	11	mention	mention	NN
zkie-zero-2020	75	12	and	and	CC
zkie-zero-2020	75	13	the	the	DT
zkie-zero-2020	75	14	sen-	sen-	NN
zkie-zero-2020	75	15	tence	tence	NN
zkie-zero-2020	75	16	the	the	DT
zkie-zero-2020	75	17	mention	mention	NN
zkie-zero-2020	75	18	is	be	VBZ
zkie-zero-2020	75	19	in	in	IN
zkie-zero-2020	75	20	the	the	DT
zkie-zero-2020	75	21	context	context	NN
zkie-zero-2020	75	22	.	.	.
zkie-zero-2020	76	1	Each	each	DT
zkie-zero-2020	76	2	candidate	candidate	NN
zkie-zero-2020	76	3	from	from	IN
zkie-zero-2020	76	4	the	the	DT
zkie-zero-2020	76	5	knowledge	knowledge	NN
zkie-zero-2020	76	6	base	base	NN
zkie-zero-2020	76	7	is	be	VBZ
zkie-zero-2020	76	8	assumed	assume	VBN
zkie-zero-2020	76	9	to	to	TO
zkie-zero-2020	76	10	have	have	VB
zkie-zero-2020	76	11	a	a	DT
zkie-zero-2020	76	12	la-	la-	JJ
zkie-zero-2020	76	13	bel	bel	NN
zkie-zero-2020	76	14	and	and	CC
zkie-zero-2020	76	15	a	a	DT
zkie-zero-2020	76	16	description	description	NN
zkie-zero-2020	76	17	.	.	.
zkie-zero-2020	77	1	For	for	IN
zkie-zero-2020	77	2	instance	instance	NN
zkie-zero-2020	77	3	,	,	,
zkie-zero-2020	77	4	in	in	IN
zkie-zero-2020	77	5	Fig	Fig	NNP
zkie-zero-2020	77	6	.	.	.
zkie-zero-2020	78	1	2	2	CD
zkie-zero-2020	78	2	,	,	,
zkie-zero-2020	78	3	one	one	CD
zkie-zero-2020	78	4	mention	mention	NN
zkie-zero-2020	78	5	is	be	VBZ
zkie-zero-2020	78	6	Dublin	Dublin	NNP
zkie-zero-2020	78	7	,	,	,
zkie-zero-2020	78	8	the	the	DT
zkie-zero-2020	78	9	context	context	NN
zkie-zero-2020	78	10	is	be	VBZ
zkie-zero-2020	78	11	Dublin	Dublin	NNP
zkie-zero-2020	78	12	is	be	VBZ
zkie-zero-2020	78	13	the	the	DT
zkie-zero-2020	78	14	cap-	cap-	JJ
zkie-zero-2020	78	15	ital	ital	NN
zkie-zero-2020	78	16	of	of	IN
zkie-zero-2020	78	17	Ireland	Ireland	NNP
zkie-zero-2020	78	18	,	,	,
zkie-zero-2020	78	19	the	the	DT
zkie-zero-2020	78	20	label	label	NN
zkie-zero-2020	78	21	of	of	IN
zkie-zero-2020	78	22	the	the	DT
zkie-zero-2020	78	23	the	the	DT
zkie-zero-2020	78	24	first	first	JJ
zkie-zero-2020	78	25	candidate	candidate	NN
zkie-zero-2020	78	26	is	be	VBZ
zkie-zero-2020	78	27	Trinity	Trinity	NNP
zkie-zero-2020	78	28	College	College	NNP
zkie-zero-2020	78	29	and	and	CC
zkie-zero-2020	78	30	its	-PRON-	PRP$
zkie-zero-2020	78	31	description	description	NN
zkie-zero-2020	78	32	is	be	VBZ
zkie-zero-2020	78	33	constituent	constituent	NN
zkie-zero-2020	78	34	college	college	NN
zkie-zero-2020	78	35	of	of	IN
zkie-zero-2020	78	36	the	the	DT
zkie-zero-2020	78	37	University	University	NNP
zkie-zero-2020	78	38	of	of	IN
zkie-zero-2020	78	39	Dublin	Dublin	NNP
zkie-zero-2020	78	40	in	in	IN
zkie-zero-2020	78	41	Ireland	Ireland	NNP
zkie-zero-2020	78	42	.	.	.
zkie-zero-2020	79	1	Mention	mention	VB
zkie-zero-2020	79	2	Detection	detection	NN
zkie-zero-2020	79	3	In	in	IN
zkie-zero-2020	79	4	the	the	DT
zkie-zero-2020	79	5	annotation	annotation	NN
zkie-zero-2020	79	6	setting	setting	NN
zkie-zero-2020	79	7	,	,	,
zkie-zero-2020	79	8	we	-PRON-	PRP
zkie-zero-2020	79	9	rely	rely	VBP
zkie-zero-2020	79	10	on	on	IN
zkie-zero-2020	79	11	users	user	NNS
zkie-zero-2020	79	12	to	to	TO
zkie-zero-2020	79	13	mark	mark	VB
zkie-zero-2020	79	14	text	text	NN
zkie-zero-2020	79	15	spans	span	NNS
zkie-zero-2020	79	16	that	that	WDT
zkie-zero-2020	79	17	contain	contain	VBP
zkie-zero-2020	79	18	annota-	annota-	JJ
zkie-zero-2020	79	19	tions	tion	NNS
zkie-zero-2020	79	20	.	.	.
zkie-zero-2020	80	1	As	as	IN
zkie-zero-2020	80	2	support	support	NN
zkie-zero-2020	80	3	,	,	,
zkie-zero-2020	80	4	we	-PRON-	PRP
zkie-zero-2020	80	5	provide	provide	VBP
zkie-zero-2020	80	6	suggestions	suggestion	NNS
zkie-zero-2020	80	7	given	give	VBN
zkie-zero-2020	80	8	by	by	IN
zkie-zero-2020	80	9	different	different	JJ
zkie-zero-2020	80	10	recommender	recommender	NN
zkie-zero-2020	80	11	models	model	NNS
zkie-zero-2020	80	12	:	:	:
zkie-zero-2020	80	13	similar	similar	JJ
zkie-zero-2020	80	14	to	to	IN
zkie-zero-2020	80	15	Yimam	Yimam	NNP
zkie-zero-2020	80	16	et	et	NNP
zkie-zero-2020	80	17	al	al	NNP
zkie-zero-2020	80	18	.	.	.
zkie-zero-2020	81	1	(	(	-LRB-
zkie-zero-2020	81	2	2014	2014	CD
zkie-zero-2020	81	3	)	)	-RRB-
zkie-zero-2020	81	4	,	,	,
zkie-zero-2020	81	5	we	-PRON-	PRP
zkie-zero-2020	81	6	use	use	VBP
zkie-zero-2020	81	7	a	a	DT
zkie-zero-2020	81	8	string	string	NN
zkie-zero-2020	81	9	matcher	matcher	NN
zkie-zero-2020	81	10	suggesting	suggest	VBG
zkie-zero-2020	81	11	an-	an-	XX
zkie-zero-2020	81	12	notations	notation	NNS
zkie-zero-2020	81	13	for	for	IN
zkie-zero-2020	81	14	mentions	mention	NNS
zkie-zero-2020	81	15	which	which	WDT
zkie-zero-2020	81	16	have	have	VBP
zkie-zero-2020	81	17	been	be	VBN
zkie-zero-2020	81	18	annotated	annotate	VBN
zkie-zero-2020	81	19	before	before	RB
zkie-zero-2020	81	20	.	.	.
zkie-zero-2020	82	1	We	-PRON-	PRP
zkie-zero-2020	82	2	also	also	RB
zkie-zero-2020	82	3	propose	propose	VBP
zkie-zero-2020	82	4	a	a	DT
zkie-zero-2020	82	5	new	new	JJ
zkie-zero-2020	82	6	Levenshtein	Levenshtein	NNP
zkie-zero-2020	82	7	string	string	NN
zkie-zero-2020	82	8	matcher	matcher	NN
zkie-zero-2020	82	9	based	base	VBN
zkie-zero-2020	82	10	on	on	IN
zkie-zero-2020	82	11	Levenshtein	Levenshtein	NNP
zkie-zero-2020	82	12	automata	automata	NN
zkie-zero-2020	82	13	(	(	-LRB-
zkie-zero-2020	82	14	Schulz	Schulz	NNP
zkie-zero-2020	82	15	and	and	CC
zkie-zero-2020	82	16	Mihov	Mihov	NNP
zkie-zero-2020	82	17	,	,	,
zkie-zero-2020	82	18	2002	2002	CD
zkie-zero-2020	82	19	)	)	-RRB-
zkie-zero-2020	82	20	.	.	.
zkie-zero-2020	83	1	In	in	IN
zkie-zero-2020	83	2	contrast	contrast	NN
zkie-zero-2020	83	3	to	to	IN
zkie-zero-2020	83	4	the	the	DT
zkie-zero-2020	83	5	string	string	NN
zkie-zero-2020	83	6	matcher	matcher	NN
zkie-zero-2020	83	7	,	,	,
zkie-zero-2020	83	8	it	-PRON-	PRP
zkie-zero-2020	83	9	suggests	suggest	VBZ
zkie-zero-2020	83	10	annotations	annotation	NNS
zkie-zero-2020	83	11	for	for	IN
zkie-zero-2020	83	12	spans	span	NNS
zkie-zero-2020	83	13	within	within	IN
zkie-zero-2020	83	14	a	a	DT
zkie-zero-2020	83	15	Leven-	Leven-	NNP
zkie-zero-2020	83	16	shtein	shtein	NN
zkie-zero-2020	83	17	distance	distance	NN
zkie-zero-2020	83	18	of	of	IN
zkie-zero-2020	83	19	1	1	CD
zkie-zero-2020	83	20	or	or	CC
zkie-zero-2020	83	21	2	2	CD
zkie-zero-2020	83	22	.	.	.
zkie-zero-2020	84	1	Preliminary	preliminary	JJ
zkie-zero-2020	84	2	experiments	experiment	NNS
zkie-zero-2020	84	3	with	with	IN
zkie-zero-2020	84	4	ML	ML	NNP
zkie-zero-2020	84	5	models	model	NNS
zkie-zero-2020	84	6	for	for	IN
zkie-zero-2020	84	7	mention	mention	NN
zkie-zero-2020	84	8	detection	detection	NN
zkie-zero-2020	84	9	like	like	IN
zkie-zero-2020	84	10	using	use	VBG
zkie-zero-2020	84	11	a	a	DT
zkie-zero-2020	84	12	Conditional	Conditional	NNP
zkie-zero-2020	84	13	Random	Random	NNP
zkie-zero-2020	84	14	Field	Field	NNP
zkie-zero-2020	84	15	and	and	CC
zkie-zero-2020	84	16	handcrafted	handcraft	VBD
zkie-zero-2020	84	17	fea-	fea-	NNS
zkie-zero-2020	84	18	tures	ture	NNS
zkie-zero-2020	84	19	did	do	VBD
zkie-zero-2020	84	20	not	not	RB
zkie-zero-2020	84	21	perform	perform	VB
zkie-zero-2020	84	22	well	well	RB
zkie-zero-2020	84	23	and	and	CC
zkie-zero-2020	84	24	yielded	yield	VBN
zkie-zero-2020	84	25	noisy	noisy	JJ
zkie-zero-2020	84	26	sug-	sug-	NN
zkie-zero-2020	84	27	gestions	gestion	NNS
zkie-zero-2020	84	28	,	,	,
zkie-zero-2020	84	29	requiring	require	VBG
zkie-zero-2020	84	30	further	further	JJ
zkie-zero-2020	84	31	investigation	investigation	NN
zkie-zero-2020	84	32	.	.	.
zkie-zero-2020	85	1	Candidate	candidate	NN
zkie-zero-2020	85	2	Generation	Generation	NNP
zkie-zero-2020	85	3	We	-PRON-	PRP
zkie-zero-2020	85	4	index	index	VBP
zkie-zero-2020	85	5	the	the	DT
zkie-zero-2020	85	6	knowledge	knowledge	NN
zkie-zero-2020	85	7	base	base	NN
zkie-zero-2020	85	8	and	and	CC
zkie-zero-2020	85	9	use	use	VB
zkie-zero-2020	85	10	full	full	JJ
zkie-zero-2020	85	11	text	text	NN
zkie-zero-2020	85	12	search	search	NN
zkie-zero-2020	85	13	to	to	TO
zkie-zero-2020	85	14	retrieve	retrieve	VB
zkie-zero-2020	85	15	candidates	candidate	NNS
zkie-zero-2020	85	16	based	base	VBN
zkie-zero-2020	85	17	on	on	IN
zkie-zero-2020	85	18	the	the	DT
zkie-zero-2020	85	19	surface	surface	NN
zkie-zero-2020	85	20	form	form	NN
zkie-zero-2020	85	21	of	of	IN
zkie-zero-2020	85	22	the	the	DT
zkie-zero-2020	85	23	annotated	annotate	VBN
zkie-zero-2020	85	24	men-	men-	NN
zkie-zero-2020	85	25	tion	tion	NN
zkie-zero-2020	85	26	.	.	.
zkie-zero-2020	86	1	Besides	besides	RB
zkie-zero-2020	86	2	,	,	,
zkie-zero-2020	86	3	users	user	NNS
zkie-zero-2020	86	4	can	can	MD
zkie-zero-2020	86	5	query	query	VB
zkie-zero-2020	86	6	this	this	DT
zkie-zero-2020	86	7	index	index	NN
zkie-zero-2020	86	8	during	during	IN
zkie-zero-2020	86	9	annotation	annotation	NN
zkie-zero-2020	86	10	.	.	.
zkie-zero-2020	87	1	We	-PRON-	PRP
zkie-zero-2020	87	2	use	use	VBP
zkie-zero-2020	87	3	fuzzy	fuzzy	JJ
zkie-zero-2020	87	4	search	search	NN
zkie-zero-2020	87	5	to	to	TO
zkie-zero-2020	87	6	help	help	VB
zkie-zero-2020	87	7	in	in	IN
zkie-zero-2020	87	8	cases	case	NNS
zkie-zero-2020	87	9	where	where	WRB
zkie-zero-2020	87	10	the	the	DT
zkie-zero-2020	87	11	mention	mention	NN
zkie-zero-2020	87	12	and	and	CC
zkie-zero-2020	87	13	the	the	DT
zkie-zero-2020	87	14	knowledge	knowledge	NN
zkie-zero-2020	87	15	base	base	NN
zkie-zero-2020	87	16	label	label	NN
zkie-zero-2020	87	17	are	be	VBP
zkie-zero-2020	87	18	almost	almost	RB
zkie-zero-2020	87	19	the	the	DT
zkie-zero-2020	87	20	same	same	JJ
zkie-zero-2020	87	21	but	but	CC
zkie-zero-2020	87	22	not	not	RB
zkie-zero-2020	87	23	identical	identical	JJ
zkie-zero-2020	87	24	(	(	-LRB-
zkie-zero-2020	87	25	e.g.	e.g.	RB
zkie-zero-2020	88	1	Dublin	Dublin	NNP
zkie-zero-2020	88	2	vs.	vs.	IN
zkie-zero-2020	88	3	Dublyn	Dublyn	NNP
zkie-zero-2020	88	4	)	)	-RRB-
zkie-zero-2020	88	5	.	.	.
zkie-zero-2020	89	1	In	in	IN
zkie-zero-2020	89	2	the	the	DT
zkie-zero-2020	89	3	interactive	interactive	JJ
zkie-zero-2020	89	4	setting	setting	NN
zkie-zero-2020	89	5	,	,	,
zkie-zero-2020	89	6	the	the	DT
zkie-zero-2020	89	7	user	user	NN
zkie-zero-2020	89	8	can	can	MD
zkie-zero-2020	89	9	also	also	RB
zkie-zero-2020	89	10	search	search	VB
zkie-zero-2020	89	11	the	the	DT
zkie-zero-2020	89	12	knowledge	knowledge	NN
zkie-zero-2020	89	13	base	base	NN
zkie-zero-2020	89	14	during	during	IN
zkie-zero-2020	89	15	annotation	annotation	NN
zkie-zero-2020	89	16	,	,	,
zkie-zero-2020	89	17	e.g.	e.g.	RB
zkie-zero-2020	90	1	in	in	IN
zkie-zero-2020	90	2	cases	case	NNS
zkie-zero-2020	90	3	when	when	WRB
zkie-zero-2020	90	4	the	the	DT
zkie-zero-2020	90	5	gold	gold	JJ
zkie-zero-2020	90	6	entity	entity	NN
zkie-zero-2020	90	7	is	be	VBZ
zkie-zero-2020	90	8	not	not	RB
zkie-zero-2020	90	9	ranked	rank	VBN
zkie-zero-2020	90	10	high	high	RB
zkie-zero-2020	90	11	enough	enough	RB
zkie-zero-2020	90	12	or	or	CC
zkie-zero-2020	90	13	when	when	WRB
zkie-zero-2020	90	14	the	the	DT
zkie-zero-2020	90	15	surface	surface	NN
zkie-zero-2020	90	16	form	form	NN
zkie-zero-2020	90	17	and	and	CC
zkie-zero-2020	90	18	knowledge	knowledge	NN
zkie-zero-2020	90	19	base	base	NN
zkie-zero-2020	90	20	label	label	NN
zkie-zero-2020	90	21	are	be	VBP
zkie-zero-2020	90	22	not	not	RB
zkie-zero-2020	90	23	the	the	DT
zkie-zero-2020	90	24	same	same	JJ
zkie-zero-2020	90	25	(	(	-LRB-
zkie-zero-2020	90	26	Zeus	Zeus	NNP
zkie-zero-2020	90	27	vs.	vs.	IN
zkie-zero-2020	90	28	Jupiter	Jupiter	NNP
zkie-zero-2020	90	29	)	)	-RRB-
zkie-zero-2020	90	30	.	.	.
zkie-zero-2020	91	1	Candidate	candidate	NN
zkie-zero-2020	91	2	Ranking	rank	VBG
zkie-zero-2020	91	3	We	-PRON-	PRP
zkie-zero-2020	91	4	follow	follow	VBP
zkie-zero-2020	91	5	Zheng	Zheng	NNP
zkie-zero-2020	91	6	et	et	FW
zkie-zero-2020	91	7	al	al	NNP
zkie-zero-2020	91	8	.	.	.
zkie-zero-2020	92	1	(	(	-LRB-
zkie-zero-2020	92	2	2010	2010	CD
zkie-zero-2020	92	3	)	)	-RRB-
zkie-zero-2020	92	4	and	and	CC
zkie-zero-2020	92	5	model	model	NN
zkie-zero-2020	92	6	candidate	candidate	NN
zkie-zero-2020	92	7	ranking	rank	VBG
zkie-zero-2020	92	8	as	as	IN
zkie-zero-2020	92	9	a	a	DT
zkie-zero-2020	92	10	learning-	learning-	NN
zkie-zero-2020	92	11	to	to	IN
zkie-zero-2020	92	12	-	-	HYPH
zkie-zero-2020	92	13	rank	rank	NN
zkie-zero-2020	92	14	problem	problem	NN
zkie-zero-2020	92	15	:	:	:
zkie-zero-2020	92	16	given	give	VBN
zkie-zero-2020	92	17	a	a	DT
zkie-zero-2020	92	18	mention	mention	NN
zkie-zero-2020	92	19	and	and	CC
zkie-zero-2020	92	20	a	a	DT
zkie-zero-2020	92	21	list	list	NN
zkie-zero-2020	92	22	of	of	IN
zkie-zero-2020	92	23	can-	can-	NN
zkie-zero-2020	92	24	didates	didate	NNS
zkie-zero-2020	92	25	,	,	,
zkie-zero-2020	92	26	sort	sort	VB
zkie-zero-2020	92	27	the	the	DT
zkie-zero-2020	92	28	candidates	candidate	NNS
zkie-zero-2020	92	29	so	so	IN
zkie-zero-2020	92	30	that	that	IN
zkie-zero-2020	92	31	the	the	DT
zkie-zero-2020	92	32	most	most	RBS
zkie-zero-2020	92	33	relevant	relevant	JJ
zkie-zero-2020	92	34	candidate	candidate	NN
zkie-zero-2020	92	35	is	be	VBZ
zkie-zero-2020	92	36	at	at	IN
zkie-zero-2020	92	37	the	the	DT
zkie-zero-2020	92	38	top	top	NN
zkie-zero-2020	92	39	.	.	.
zkie-zero-2020	93	1	For	for	IN
zkie-zero-2020	93	2	training	training	NN
zkie-zero-2020	93	3	,	,	,
zkie-zero-2020	93	4	we	-PRON-	PRP
zkie-zero-2020	93	5	guarantee	guarantee	VBP
zkie-zero-2020	93	6	that	that	IN
zkie-zero-2020	93	7	the	the	DT
zkie-zero-2020	93	8	gold	gold	JJ
zkie-zero-2020	93	9	candidate	candidate	NN
zkie-zero-2020	93	10	is	be	VBZ
zkie-zero-2020	93	11	present	present	JJ
zkie-zero-2020	93	12	in	in	IN
zkie-zero-2020	93	13	the	the	DT
zkie-zero-2020	93	14	candidate	candidate	NN
zkie-zero-2020	93	15	list	list	NN
zkie-zero-2020	93	16	.	.	.
zkie-zero-2020	94	1	For	for	IN
zkie-zero-2020	94	2	evaluation	evaluation	NN
zkie-zero-2020	94	3	,	,	,
zkie-zero-2020	94	4	the	the	DT
zkie-zero-2020	94	5	gold	gold	NN
zkie-zero-2020	94	6	candidate	candidate	NN
zkie-zero-2020	94	7	can	can	MD
zkie-zero-2020	94	8	be	be	VB
zkie-zero-2020	94	9	ab-	ab-	RB
zkie-zero-2020	94	10	sent	send	VBN
zkie-zero-2020	94	11	from	from	IN
zkie-zero-2020	94	12	the	the	DT
zkie-zero-2020	94	13	candidate	candidate	NN
zkie-zero-2020	94	14	list	list	NN
zkie-zero-2020	94	15	if	if	IN
zkie-zero-2020	94	16	the	the	DT
zkie-zero-2020	94	17	candidate	candidate	NN
zkie-zero-2020	94	18	search	search	NN
zkie-zero-2020	94	19	failed	fail	VBD
zkie-zero-2020	94	20	to	to	TO
zkie-zero-2020	94	21	find	find	VB
zkie-zero-2020	94	22	it	-PRON-	PRP
zkie-zero-2020	94	23	.	.	.
zkie-zero-2020	95	1	This	this	DT
zkie-zero-2020	95	2	interaction	interaction	NN
zkie-zero-2020	95	3	is	be	VBZ
zkie-zero-2020	95	4	the	the	DT
zkie-zero-2020	95	5	core	core	JJ
zkie-zero-2020	95	6	Human	human	NN
zkie-zero-2020	95	7	-	-	HYPH
zkie-zero-2020	95	8	in	in	IN
zkie-zero-2020	95	9	-	-	HYPH
zkie-zero-2020	95	10	the	the	DT
zkie-zero-2020	95	11	-	-	HYPH
zkie-zero-2020	95	12	loop	loop	NN
zkie-zero-2020	95	13	in	in	IN
zkie-zero-2020	95	14	our	-PRON-	PRP$
zkie-zero-2020	95	15	approach	approach	NN
zkie-zero-2020	95	16	.	.	.
zkie-zero-2020	96	1	For	for	IN
zkie-zero-2020	96	2	training	training	NN
zkie-zero-2020	96	3	,	,	,
zkie-zero-2020	96	4	we	-PRON-	PRP
zkie-zero-2020	96	5	rephrase	rephrase	VBP
zkie-zero-2020	96	6	the	the	DT
zkie-zero-2020	96	7	task	task	NN
zkie-zero-2020	96	8	as	as	IN
zkie-zero-2020	96	9	preference	preference	NN
zkie-zero-2020	96	10	learning	learn	VBG
zkie-zero-2020	96	11	:	:	:
zkie-zero-2020	96	12	By	by	IN
zkie-zero-2020	96	13	selecting	select	VBG
zkie-zero-2020	96	14	an	an	DT
zkie-zero-2020	96	15	entity	entity	NN
zkie-zero-2020	96	16	label	label	NN
zkie-zero-2020	96	17	from	from	IN
zkie-zero-2020	96	18	the	the	DT
zkie-zero-2020	96	19	candidate	candidate	NN
zkie-zero-2020	96	20	list	list	NN
zkie-zero-2020	96	21	,	,	,
zkie-zero-2020	96	22	users	user	NNS
zkie-zero-2020	96	23	express	express	VBP
zkie-zero-2020	96	24	that	that	IN
zkie-zero-2020	96	25	the	the	DT
zkie-zero-2020	96	26	se-	se-	NN
zkie-zero-2020	96	27	lected	lecte	VBD
zkie-zero-2020	96	28	one	one	CD
zkie-zero-2020	96	29	was	be	VBD
zkie-zero-2020	96	30	preferred	prefer	VBN
zkie-zero-2020	96	31	over	over	IN
zkie-zero-2020	96	32	all	all	DT
zkie-zero-2020	96	33	other	other	JJ
zkie-zero-2020	96	34	candidates	candidate	NNS
zkie-zero-2020	96	35	.	.	.
zkie-zero-2020	97	1	These	these	DT
zkie-zero-2020	97	2	preferences	preference	NNS
zkie-zero-2020	97	3	are	be	VBP
zkie-zero-2020	97	4	used	use	VBN
zkie-zero-2020	97	5	to	to	TO
zkie-zero-2020	97	6	train	train	VB
zkie-zero-2020	97	7	state	state	NN
zkie-zero-2020	97	8	-	-	HYPH
zkie-zero-2020	97	9	of	of	IN
zkie-zero-2020	97	10	-	-	HYPH
zkie-zero-2020	97	11	the	the	DT
zkie-zero-2020	97	12	-	-	HYPH
zkie-zero-2020	97	13	art	art	NN
zkie-zero-2020	97	14	pairwise	pairwise	NN
zkie-zero-2020	97	15	learning	learn	VBG
zkie-zero-2020	97	16	-	-	HYPH
zkie-zero-2020	97	17	to	to	IN
zkie-zero-2020	97	18	-	-	HYPH
zkie-zero-2020	97	19	rank	rank	NN
zkie-zero-2020	97	20	models	model	NNS
zkie-zero-2020	97	21	from	from	IN
zkie-zero-2020	97	22	the	the	DT
zkie-zero-2020	97	23	litera-	litera-	NNP
zkie-zero-2020	97	24	ture	ture	NN
zkie-zero-2020	97	25	:	:	:
zkie-zero-2020	97	26	the	the	DT
zkie-zero-2020	97	27	gradient	gradient	NN
zkie-zero-2020	97	28	boosted	boost	VBD
zkie-zero-2020	97	29	trees	tree	NNS
zkie-zero-2020	97	30	variant	variant	JJ
zkie-zero-2020	97	31	LightGBM	LightGBM	.
zkie-zero-2020	97	32	(	(	-LRB-
zkie-zero-2020	97	33	Ke	Ke	NNP
zkie-zero-2020	97	34	et	et	FW
zkie-zero-2020	97	35	al	al	NNP
zkie-zero-2020	97	36	.	.	NNP
zkie-zero-2020	97	37	,	,	,
zkie-zero-2020	97	38	2017	2017	CD
zkie-zero-2020	97	39	)	)	-RRB-
zkie-zero-2020	97	40	,	,	,
zkie-zero-2020	97	41	RankSVM	RankSVM	.
zkie-zero-2020	97	42	(	(	-LRB-
zkie-zero-2020	97	43	Joachims	Joachims	NNP
zkie-zero-2020	97	44	,	,	,
zkie-zero-2020	97	45	2002	2002	CD
zkie-zero-2020	97	46	)	)	-RRB-
zkie-zero-2020	97	47	and	and	CC
zkie-zero-2020	97	48	RankNet	RankNet	NNP
zkie-zero-2020	97	49	(	(	-LRB-
zkie-zero-2020	97	50	Burges	Burges	NNP
zkie-zero-2020	97	51	et	et	NNP
zkie-zero-2020	97	52	al	al	NNP
zkie-zero-2020	97	53	.	.	NNP
zkie-zero-2020	97	54	,	,	,
zkie-zero-2020	97	55	2005	2005	CD
zkie-zero-2020	97	56	)	)	-RRB-
zkie-zero-2020	97	57	.	.	.
zkie-zero-2020	98	1	Models	model	NNS
zkie-zero-2020	98	2	are	be	VBP
zkie-zero-2020	98	3	re-	re-	RB
zkie-zero-2020	98	4	trained	train	VBN
zkie-zero-2020	98	5	in	in	IN
zkie-zero-2020	98	6	the	the	DT
zkie-zero-2020	98	7	background	background	NN
zkie-zero-2020	98	8	when	when	WRB
zkie-zero-2020	98	9	new	new	JJ
zkie-zero-2020	98	10	annotations	annotation	NNS
zkie-zero-2020	98	11	are	be	VBP
zkie-zero-2020	98	12	made	make	VBN
zkie-zero-2020	98	13	,	,	,
zkie-zero-2020	98	14	thus	thus	RB
zkie-zero-2020	98	15	improving	improve	VBG
zkie-zero-2020	98	16	over	over	IN
zkie-zero-2020	98	17	time	time	NN
zkie-zero-2020	98	18	with	with	IN
zkie-zero-2020	98	19	an	an	DT
zkie-zero-2020	98	20	in-	in-	JJ
zkie-zero-2020	98	21	creasing	creasing	NN
zkie-zero-2020	98	22	number	number	NN
zkie-zero-2020	98	23	of	of	IN
zkie-zero-2020	98	24	annotations	annotation	NNS
zkie-zero-2020	98	25	.	.	.
zkie-zero-2020	99	1	We	-PRON-	PRP
zkie-zero-2020	99	2	use	use	VBP
zkie-zero-2020	99	3	a	a	DT
zkie-zero-2020	99	4	set	set	NN
zkie-zero-2020	99	5	of	of	IN
zkie-zero-2020	99	6	generic	generic	JJ
zkie-zero-2020	99	7	handcrafted	handcraft	VBN
zkie-zero-2020	99	8	features	feature	NNS
zkie-zero-2020	99	9	which	which	WDT
zkie-zero-2020	99	10	are	be	VBP
zkie-zero-2020	99	11	described	describe	VBN
zkie-zero-2020	99	12	in	in	IN
zkie-zero-2020	99	13	Table	table	NN
zkie-zero-2020	99	14	1	1	CD
zkie-zero-2020	99	15	.	.	.
zkie-zero-2020	100	1	These	these	DT
zkie-zero-2020	100	2	models	model	NNS
zkie-zero-2020	100	3	were	be	VBD
zkie-zero-2020	100	4	chosen	choose	VBN
zkie-zero-2020	100	5	as	as	IN
zkie-zero-2020	100	6	they	-PRON-	PRP
zkie-zero-2020	100	7	can	can	MD
zkie-zero-2020	100	8	work	work	VB
zkie-zero-2020	100	9	with	with	IN
zkie-zero-2020	100	10	low	low	JJ
zkie-zero-2020	100	11	data	datum	NNS
zkie-zero-2020	100	12	,	,	,
zkie-zero-2020	100	13	train	train	VB
zkie-zero-2020	100	14	quickly	quickly	RB
zkie-zero-2020	100	15	and	and	CC
zkie-zero-2020	100	16	allow	allow	VB
zkie-zero-2020	100	17	intro-	intro-	NN
zkie-zero-2020	100	18	spection	spection	NN
zkie-zero-2020	100	19	.	.	.
zkie-zero-2020	101	1	Using	use	VBG
zkie-zero-2020	101	2	deep	deep	JJ
zkie-zero-2020	101	3	models	model	NNS
zkie-zero-2020	101	4	or	or	CC
zkie-zero-2020	101	5	word	word	NN
zkie-zero-2020	101	6	embeddings	embedding	NNS
zkie-zero-2020	101	7	as	as	IN
zkie-zero-2020	101	8	input	input	NN
zkie-zero-2020	101	9	features	feature	NNS
zkie-zero-2020	101	10	showed	show	VBD
zkie-zero-2020	101	11	to	to	TO
zkie-zero-2020	101	12	be	be	VB
zkie-zero-2020	101	13	too	too	RB
zkie-zero-2020	101	14	slow	slow	JJ
zkie-zero-2020	101	15	to	to	TO
zkie-zero-2020	101	16	be	be	VB
zkie-zero-2020	101	17	inter-	inter-	VBG
zkie-zero-2020	101	18	6985	6985	CD
zkie-zero-2020	101	19	active	active	JJ
zkie-zero-2020	101	20	.	.	.
zkie-zero-2020	102	1	We	-PRON-	PRP
zkie-zero-2020	102	2	also	also	RB
zkie-zero-2020	102	3	leverage	leverage	VBP
zkie-zero-2020	102	4	pretrained	pretraine	VBN
zkie-zero-2020	102	5	Sentence	Sentence	NNP
zkie-zero-2020	102	6	-	-	HYPH
zkie-zero-2020	102	7	BERT	BERT	NNP
zkie-zero-2020	102	8	embeddings	embedding	NNS
zkie-zero-2020	102	9	(	(	-LRB-
zkie-zero-2020	102	10	Reimers	Reimers	NNPS
zkie-zero-2020	102	11	and	and	CC
zkie-zero-2020	102	12	Gurevych	Gurevych	NNP
zkie-zero-2020	102	13	,	,	,
zkie-zero-2020	102	14	2019	2019	CD
zkie-zero-2020	102	15	)	)	-RRB-
zkie-zero-2020	102	16	trained	train	VBN
zkie-zero-2020	102	17	on	on	IN
zkie-zero-2020	102	18	Natural	Natural	NNP
zkie-zero-2020	102	19	Language	Language	NNP
zkie-zero-2020	102	20	Inference	Inference	NNP
zkie-zero-2020	102	21	data	datum	NNS
zkie-zero-2020	102	22	written	write	VBN
zkie-zero-2020	102	23	in	in	IN
zkie-zero-2020	102	24	sim-	sim-	NNP
zkie-zero-2020	102	25	ple	ple	NNP
zkie-zero-2020	102	26	English	English	NNP
zkie-zero-2020	102	27	.	.	.
zkie-zero-2020	103	1	These	these	DT
zkie-zero-2020	103	2	are	be	VBP
zkie-zero-2020	103	3	not	not	RB
zkie-zero-2020	103	4	fine	fine	RB
zkie-zero-2020	103	5	-	-	HYPH
zkie-zero-2020	103	6	tuned	tune	VBN
zkie-zero-2020	103	7	by	by	IN
zkie-zero-2020	103	8	us	-PRON-	PRP
zkie-zero-2020	103	9	during	during	IN
zkie-zero-2020	103	10	training	training	NN
zkie-zero-2020	103	11	.	.	.
zkie-zero-2020	104	1	Although	although	IN
zkie-zero-2020	104	2	they	-PRON-	PRP
zkie-zero-2020	104	3	come	come	VBP
zkie-zero-2020	104	4	from	from	IN
zkie-zero-2020	104	5	a	a	DT
zkie-zero-2020	104	6	different	different	JJ
zkie-zero-2020	104	7	do-	do-	JJ
zkie-zero-2020	104	8	main	main	NN
zkie-zero-2020	104	9	,	,	,
zkie-zero-2020	104	10	we	-PRON-	PRP
zkie-zero-2020	104	11	conjecture	conjecture	VBP
zkie-zero-2020	104	12	that	that	IN
zkie-zero-2020	104	13	the	the	DT
zkie-zero-2020	104	14	WordPiece	WordPiece	NNP
zkie-zero-2020	104	15	tokeniza-	tokeniza-	NN
zkie-zero-2020	104	16	tion	tion	NN
zkie-zero-2020	104	17	of	of	IN
zkie-zero-2020	104	18	BERT	BERT	NNP
zkie-zero-2020	104	19	helps	help	VBZ
zkie-zero-2020	104	20	with	with	IN
zkie-zero-2020	104	21	the	the	DT
zkie-zero-2020	104	22	spelling	spelling	NN
zkie-zero-2020	104	23	variance	variance	NN
zkie-zero-2020	104	24	of	of	IN
zkie-zero-2020	104	25	our	-PRON-	PRP$
zkie-zero-2020	104	26	texts	text	NNS
zkie-zero-2020	104	27	in	in	IN
zkie-zero-2020	104	28	contrast	contrast	NN
zkie-zero-2020	104	29	to	to	IN
zkie-zero-2020	104	30	traditional	traditional	JJ
zkie-zero-2020	104	31	word	word	NN
zkie-zero-2020	104	32	embeddings	embedding	NNS
zkie-zero-2020	104	33	which	which	WDT
zkie-zero-2020	104	34	would	would	MD
zkie-zero-2020	104	35	have	have	VB
zkie-zero-2020	104	36	many	many	JJ
zkie-zero-2020	104	37	out	out	JJ
zkie-zero-2020	104	38	-	-	HYPH
zkie-zero-2020	104	39	of	of	IN
zkie-zero-2020	104	40	-	-	HYPH
zkie-zero-2020	104	41	vocabulary	vocabulary	NN
zkie-zero-2020	104	42	words	word	NNS
zkie-zero-2020	104	43	.	.	.
zkie-zero-2020	105	1	For	for	IN
zkie-zero-2020	105	2	specific	specific	JJ
zkie-zero-2020	105	3	tasks	task	NNS
zkie-zero-2020	105	4	,	,	,
zkie-zero-2020	105	5	custom	custom	NN
zkie-zero-2020	105	6	features	feature	NNS
zkie-zero-2020	105	7	can	can	MD
zkie-zero-2020	105	8	easily	easily	RB
zkie-zero-2020	105	9	be	be	VB
zkie-zero-2020	105	10	incorporated	incorporate	VBN
zkie-zero-2020	105	11	e.g.	e.g.	RB
zkie-zero-2020	106	1	entity	entity	NN
zkie-zero-2020	106	2	type	type	NN
zkie-zero-2020	106	3	information	information	NN
zkie-zero-2020	106	4	,	,	,
zkie-zero-2020	106	5	time	time	NN
zkie-zero-2020	106	6	in-	in-	JJ
zkie-zero-2020	106	7	formation	formation	NN
zkie-zero-2020	106	8	for	for	IN
zkie-zero-2020	106	9	diachronic	diachronic	JJ
zkie-zero-2020	106	10	entity	entity	NN
zkie-zero-2020	106	11	linking	linking	NN
zkie-zero-2020	106	12	,	,	,
zkie-zero-2020	106	13	location	location	NN
zkie-zero-2020	106	14	information	information	NN
zkie-zero-2020	106	15	or	or	CC
zkie-zero-2020	106	16	distance	distance	NN
zkie-zero-2020	106	17	for	for	IN
zkie-zero-2020	106	18	annotating	annotate	VBG
zkie-zero-2020	106	19	geographi-	geographi-	NN
zkie-zero-2020	106	20	cal	cal	NN
zkie-zero-2020	106	21	entities	entity	NNS
zkie-zero-2020	106	22	.	.	.
zkie-zero-2020	107	1	•	•	NNP
zkie-zero-2020	107	2	Mention	mention	VBP
zkie-zero-2020	107	3	exactly	exactly	RB
zkie-zero-2020	107	4	matches	match	VBZ
zkie-zero-2020	107	5	label	label	NN
zkie-zero-2020	107	6	•	•	NNP
zkie-zero-2020	107	7	Label	Label	NNP
zkie-zero-2020	107	8	is	be	VBZ
zkie-zero-2020	107	9	prefix	prefix	NN
zkie-zero-2020	107	10	/	/	SYM
zkie-zero-2020	107	11	postfix	postfix	NN
zkie-zero-2020	107	12	of	of	IN
zkie-zero-2020	107	13	mention	mention	NN
zkie-zero-2020	107	14	•	•	NNP
zkie-zero-2020	107	15	Mention	Mention	NNP
zkie-zero-2020	107	16	is	be	VBZ
zkie-zero-2020	107	17	prefix	prefix	NN
zkie-zero-2020	107	18	/	/	SYM
zkie-zero-2020	107	19	postfix	postfix	NN
zkie-zero-2020	107	20	of	of	IN
zkie-zero-2020	107	21	label	label	NNP
zkie-zero-2020	107	22	•	•	NNP
zkie-zero-2020	107	23	Label	Label	NNP
zkie-zero-2020	107	24	is	be	VBZ
zkie-zero-2020	107	25	substring	substre	VBG
zkie-zero-2020	107	26	of	of	IN
zkie-zero-2020	107	27	mention	mention	NN
zkie-zero-2020	107	28	and	and	CC
zkie-zero-2020	107	29	vice	vice	NN
zkie-zero-2020	107	30	versa	versa	NN
zkie-zero-2020	107	31	•	•	NNP
zkie-zero-2020	107	32	Levenshtein	Levenshtein	NNP
zkie-zero-2020	107	33	distance	distance	NN
zkie-zero-2020	107	34	between	between	IN
zkie-zero-2020	107	35	mention	mention	NN
zkie-zero-2020	107	36	and	and	CC
zkie-zero-2020	107	37	label	label	NNP
zkie-zero-2020	107	38	•	•	NNP
zkie-zero-2020	107	39	Levenshtein	Levenshtein	NNP
zkie-zero-2020	107	40	distance	distance	NN
zkie-zero-2020	107	41	between	between	IN
zkie-zero-2020	107	42	context	context	NN
zkie-zero-2020	107	43	and	and	CC
zkie-zero-2020	107	44	description	description	NN
zkie-zero-2020	107	45	•	•	NNP
zkie-zero-2020	107	46	Jaro	Jaro	NNP
zkie-zero-2020	107	47	-	-	HYPH
zkie-zero-2020	107	48	Winkler	Winkler	NNP
zkie-zero-2020	107	49	distance	distance	NN
zkie-zero-2020	107	50	between	between	IN
zkie-zero-2020	107	51	mention	mention	NN
zkie-zero-2020	107	52	and	and	CC
zkie-zero-2020	107	53	label	label	NN
zkie-zero-2020	107	54	•	•	NNP
zkie-zero-2020	107	55	Jaro	Jaro	NNP
zkie-zero-2020	107	56	-	-	HYPH
zkie-zero-2020	107	57	Winkler	Winkler	NNP
zkie-zero-2020	107	58	distance	distance	NN
zkie-zero-2020	107	59	between	between	IN
zkie-zero-2020	107	60	context	context	NN
zkie-zero-2020	107	61	and	and	CC
zkie-zero-2020	107	62	description	description	NN
zkie-zero-2020	107	63	•	•	NNP
zkie-zero-2020	107	64	Sørensen	Sørensen	NNP
zkie-zero-2020	107	65	-	-	HYPH
zkie-zero-2020	107	66	Dice	Dice	NNP
zkie-zero-2020	107	67	index	index	NN
zkie-zero-2020	107	68	between	between	IN
zkie-zero-2020	107	69	context	context	NN
zkie-zero-2020	107	70	and	and	CC
zkie-zero-2020	107	71	description	description	NN
zkie-zero-2020	107	72	•	•	VBP
zkie-zero-2020	107	73	Jaccard	Jaccard	NNP
zkie-zero-2020	107	74	coefficient	coefficient	NN
zkie-zero-2020	107	75	between	between	IN
zkie-zero-2020	107	76	context	context	NN
zkie-zero-2020	107	77	and	and	CC
zkie-zero-2020	107	78	description	description	NN
zkie-zero-2020	107	79	•	•	VBP
zkie-zero-2020	107	80	Exact	exact	JJ
zkie-zero-2020	107	81	match	match	NN
zkie-zero-2020	107	82	of	of	IN
zkie-zero-2020	107	83	Soundex	Soundex	NNP
zkie-zero-2020	107	84	encoding	encoding	NN
zkie-zero-2020	107	85	of	of	IN
zkie-zero-2020	107	86	mention	mention	NN
zkie-zero-2020	107	87	and	and	CC
zkie-zero-2020	107	88	label	label	NN
zkie-zero-2020	107	89	•	•	NNP
zkie-zero-2020	107	90	Phonetic	Phonetic	NNP
zkie-zero-2020	107	91	Match	Match	NNP
zkie-zero-2020	107	92	Rating	rating	NN
zkie-zero-2020	107	93	of	of	IN
zkie-zero-2020	107	94	mention	mention	NN
zkie-zero-2020	107	95	and	and	CC
zkie-zero-2020	107	96	label	label	NN
zkie-zero-2020	107	97	•	•	NNP
zkie-zero-2020	107	98	Cosine	Cosine	NNP
zkie-zero-2020	107	99	distance	distance	NN
zkie-zero-2020	107	100	between	between	IN
zkie-zero-2020	107	101	SBERT	SBERT	NNP
zkie-zero-2020	107	102	Embeddings	embedding	NNS
zkie-zero-2020	107	103	of	of	IN
zkie-zero-2020	107	104	context	context	NN
zkie-zero-2020	107	105	and	and	CC
zkie-zero-2020	107	106	description	description	NN
zkie-zero-2020	107	107	(	(	-LRB-
zkie-zero-2020	107	108	Reimers	Reimers	NNPS
zkie-zero-2020	107	109	and	and	CC
zkie-zero-2020	107	110	Gurevych	Gurevych	NNP
zkie-zero-2020	107	111	,	,	,
zkie-zero-2020	107	112	2019	2019	CD
zkie-zero-2020	107	113	)	)	-RRB-
zkie-zero-2020	107	114	•	•	NN
zkie-zero-2020	107	115	Query	query	NN
zkie-zero-2020	107	116	length	length	NN
zkie-zero-2020	107	117	*	*	NN
zkie-zero-2020	107	118	Query	Query	NNP
zkie-zero-2020	107	119	exactly	exactly	RB
zkie-zero-2020	107	120	matches	match	VBZ
zkie-zero-2020	107	121	label	label	NN
zkie-zero-2020	107	122	*	*	NFP
zkie-zero-2020	107	123	Query	Query	NNP
zkie-zero-2020	107	124	is	be	VBZ
zkie-zero-2020	107	125	prefix	prefix	NN
zkie-zero-2020	107	126	/	/	SYM
zkie-zero-2020	107	127	postfix	postfix	NN
zkie-zero-2020	107	128	of	of	IN
zkie-zero-2020	107	129	label	label	NN
zkie-zero-2020	107	130	/	/	SYM
zkie-zero-2020	107	131	mention	mention	NN
zkie-zero-2020	107	132	*	*	NFP
zkie-zero-2020	107	133	Query	Query	NNP
zkie-zero-2020	107	134	is	be	VBZ
zkie-zero-2020	107	135	substring	substre	VBG
zkie-zero-2020	107	136	of	of	IN
zkie-zero-2020	107	137	mention	mention	NN
zkie-zero-2020	107	138	/	/	SYM
zkie-zero-2020	107	139	label	label	NN
zkie-zero-2020	107	140	*	*	NFP
zkie-zero-2020	107	141	Levenshtein	Levenshtein	NNP
zkie-zero-2020	107	142	distance	distance	NN
zkie-zero-2020	107	143	between	between	IN
zkie-zero-2020	107	144	query	query	NN
zkie-zero-2020	107	145	and	and	CC
zkie-zero-2020	107	146	label	label	NN
zkie-zero-2020	107	147	•	•	NNP
zkie-zero-2020	107	148	Levenshtein	Levenshtein	NNP
zkie-zero-2020	107	149	distance	distance	NN
zkie-zero-2020	107	150	between	between	IN
zkie-zero-2020	107	151	query	query	NN
zkie-zero-2020	107	152	and	and	CC
zkie-zero-2020	107	153	mention	mention	VB
zkie-zero-2020	107	154	•	•	NNP
zkie-zero-2020	107	155	Jaro	Jaro	NNP
zkie-zero-2020	107	156	-	-	HYPH
zkie-zero-2020	107	157	Winkler	Winkler	NNP
zkie-zero-2020	107	158	distance	distance	NN
zkie-zero-2020	107	159	between	between	IN
zkie-zero-2020	107	160	query	query	NN
zkie-zero-2020	107	161	and	and	CC
zkie-zero-2020	107	162	label	label	NN
zkie-zero-2020	107	163	•	•	NNP
zkie-zero-2020	107	164	Jaro	Jaro	NNP
zkie-zero-2020	107	165	-	-	HYPH
zkie-zero-2020	107	166	Winkler	Winkler	NNP
zkie-zero-2020	107	167	distance	distance	NN
zkie-zero-2020	107	168	between	between	IN
zkie-zero-2020	107	169	query	query	NN
zkie-zero-2020	107	170	and	and	CC
zkie-zero-2020	107	171	mention	mention	VB
zkie-zero-2020	107	172	Table	table	NN
zkie-zero-2020	107	173	1	1	CD
zkie-zero-2020	107	174	:	:	:
zkie-zero-2020	107	175	Features	feature	NNS
zkie-zero-2020	107	176	used	use	VBN
zkie-zero-2020	107	177	for	for	IN
zkie-zero-2020	107	178	candidate	candidate	NN
zkie-zero-2020	107	179	ranking	ranking	NN
zkie-zero-2020	107	180	.	.	.
zkie-zero-2020	108	1	Starred	star	VBN
zkie-zero-2020	108	2	features	feature	NNS
zkie-zero-2020	108	3	were	be	VBD
zkie-zero-2020	108	4	also	also	RB
zkie-zero-2020	108	5	used	use	VBN
zkie-zero-2020	108	6	by	by	IN
zkie-zero-2020	108	7	Zheng	Zheng	NNP
zkie-zero-2020	108	8	et	et	FW
zkie-zero-2020	108	9	al	al	NNP
zkie-zero-2020	108	10	.	.	.
zkie-zero-2020	109	1	(	(	-LRB-
zkie-zero-2020	109	2	2010	2010	CD
zkie-zero-2020	109	3	)	)	-RRB-
zkie-zero-2020	109	4	4	4	CD
zkie-zero-2020	109	5	Datasets	dataset	NNS
zkie-zero-2020	109	6	There	there	EX
zkie-zero-2020	109	7	are	be	VBP
zkie-zero-2020	109	8	very	very	RB
zkie-zero-2020	109	9	few	few	JJ
zkie-zero-2020	109	10	datasets	dataset	NNS
zkie-zero-2020	109	11	available	available	JJ
zkie-zero-2020	109	12	that	that	WDT
zkie-zero-2020	109	13	can	can	MD
zkie-zero-2020	109	14	be	be	VB
zkie-zero-2020	109	15	used	use	VBN
zkie-zero-2020	109	16	for	for	IN
zkie-zero-2020	109	17	EL	el	NN
zkie-zero-2020	109	18	against	against	IN
zkie-zero-2020	109	19	domain	domain	NN
zkie-zero-2020	109	20	-	-	HYPH
zkie-zero-2020	109	21	specific	specific	JJ
zkie-zero-2020	109	22	knowledge	knowledge	NN
zkie-zero-2020	109	23	bases	basis	NNS
zkie-zero-2020	109	24	,	,	,
zkie-zero-2020	109	25	further	further	RB
zkie-zero-2020	109	26	stressing	stress	VBG
zkie-zero-2020	109	27	our	-PRON-	PRP$
zkie-zero-2020	109	28	point	point	NN
zkie-zero-2020	109	29	that	that	IN
zkie-zero-2020	109	30	we	-PRON-	PRP
zkie-zero-2020	109	31	need	need	VBP
zkie-zero-2020	109	32	more	more	JJR
zkie-zero-2020	109	33	of	of	IN
zkie-zero-2020	109	34	these	these	DT
zkie-zero-2020	109	35	,	,	,
zkie-zero-2020	109	36	thereby	thereby	RB
zkie-zero-2020	109	37	requiring	require	VBG
zkie-zero-2020	109	38	approaches	approach	NNS
zkie-zero-2020	109	39	like	like	IN
zkie-zero-2020	109	40	ours	our	NNS
zkie-zero-2020	109	41	to	to	TO
zkie-zero-2020	109	42	create	create	VB
zkie-zero-2020	109	43	them	-PRON-	PRP
zkie-zero-2020	109	44	.	.	.
zkie-zero-2020	110	1	We	-PRON-	PRP
zkie-zero-2020	110	2	use	use	VBP
zkie-zero-2020	110	3	three	three	CD
zkie-zero-2020	110	4	datasets	dataset	NNS
zkie-zero-2020	110	5	:	:	:
zkie-zero-2020	110	6	AIDA	AIDA	NNP
zkie-zero-2020	110	7	-	-	HYPH
zkie-zero-2020	110	8	YAGO	YAGO	NNP
zkie-zero-2020	110	9	,	,	,
zkie-zero-2020	110	10	Women	Women	NNPS
zkie-zero-2020	110	11	Writers	Writers	NNPS
zkie-zero-2020	110	12	Online	Online	NNP
zkie-zero-2020	110	13	(	(	-LRB-
zkie-zero-2020	110	14	WWO	WWO	NNP
zkie-zero-2020	110	15	)	)	-RRB-
zkie-zero-2020	110	16	and	and	CC
zkie-zero-2020	110	17	1641	1641	CD
zkie-zero-2020	110	18	Deposi-	Deposi-	NNP
zkie-zero-2020	110	19	tions	tion	NNS
zkie-zero-2020	110	20	.	.	.
zkie-zero-2020	111	1	AIDA	AIDA	NNP
zkie-zero-2020	111	2	consists	consist	VBZ
zkie-zero-2020	111	3	of	of	IN
zkie-zero-2020	111	4	Reuters	Reuters	NNP
zkie-zero-2020	111	5	news	news	NN
zkie-zero-2020	111	6	stories	story	NNS
zkie-zero-2020	111	7	.	.	.
zkie-zero-2020	112	1	To	to	IN
zkie-zero-2020	112	2	the	the	DT
zkie-zero-2020	112	3	best	good	JJS
zkie-zero-2020	112	4	of	of	IN
zkie-zero-2020	112	5	our	-PRON-	PRP$
zkie-zero-2020	112	6	knowledge	knowledge	NN
zkie-zero-2020	112	7	,	,	,
zkie-zero-2020	112	8	WWO	WWO	NNP
zkie-zero-2020	112	9	has	have	VBZ
zkie-zero-2020	112	10	not	not	RB
zkie-zero-2020	112	11	been	be	VBN
zkie-zero-2020	112	12	consid-	consid-	NN
zkie-zero-2020	112	13	ered	ered	JJ
zkie-zero-2020	112	14	for	for	IN
zkie-zero-2020	112	15	automatic	automatic	JJ
zkie-zero-2020	112	16	EL	el	NN
zkie-zero-2020	112	17	so	so	RB
zkie-zero-2020	112	18	far	far	RB
zkie-zero-2020	112	19	.	.	.
zkie-zero-2020	113	1	The	the	DT
zkie-zero-2020	113	2	1641	1641	CD
zkie-zero-2020	113	3	Depositions	deposition	NNS
zkie-zero-2020	113	4	have	have	VBP
zkie-zero-2020	113	5	been	be	VBN
zkie-zero-2020	113	6	used	use	VBN
zkie-zero-2020	113	7	in	in	IN
zkie-zero-2020	113	8	automatic	automatic	JJ
zkie-zero-2020	113	9	EL	el	NN
zkie-zero-2020	113	10	,	,	,
zkie-zero-2020	113	11	but	but	CC
zkie-zero-2020	113	12	only	only	RB
zkie-zero-2020	113	13	when	when	WRB
zkie-zero-2020	113	14	linking	link	VBG
zkie-zero-2020	113	15	against	against	IN
zkie-zero-2020	113	16	DBPedia	DBPedia	NNP
zkie-zero-2020	113	17	which	which	WDT
zkie-zero-2020	113	18	has	have	VBZ
zkie-zero-2020	113	19	a	a	DT
zkie-zero-2020	113	20	very	very	RB
zkie-zero-2020	113	21	low	low	JJ
zkie-zero-2020	113	22	en-	en-	JJ
zkie-zero-2020	113	23	tity	tity	NN
zkie-zero-2020	113	24	coverage	coverage	NN
zkie-zero-2020	113	25	(	(	-LRB-
zkie-zero-2020	113	26	Munnelly	munnelly	RB
zkie-zero-2020	113	27	and	and	CC
zkie-zero-2020	113	28	Lawless	Lawless	NNP
zkie-zero-2020	113	29	,	,	,
zkie-zero-2020	113	30	2018b	2018b	CD
zkie-zero-2020	113	31	)	)	-RRB-
zkie-zero-2020	113	32	.	.	.
zkie-zero-2020	114	1	We	-PRON-	PRP
zkie-zero-2020	114	2	preprocess	preprocess	VBP
zkie-zero-2020	114	3	the	the	DT
zkie-zero-2020	114	4	data	datum	NNS
zkie-zero-2020	114	5	,	,	,
zkie-zero-2020	114	6	split	split	VBD
zkie-zero-2020	114	7	it	-PRON-	PRP
zkie-zero-2020	114	8	in	in	IN
zkie-zero-2020	114	9	sentences	sentence	NNS
zkie-zero-2020	114	10	,	,	,
zkie-zero-2020	114	11	tokenize	tokenize	VB
zkie-zero-2020	114	12	and	and	CC
zkie-zero-2020	114	13	reduce	reduce	VB
zkie-zero-2020	114	14	noise	noise	NN
zkie-zero-2020	114	15	.	.	.
zkie-zero-2020	115	1	For	for	IN
zkie-zero-2020	115	2	WWO	WWO	NNP
zkie-zero-2020	115	3	,	,	,
zkie-zero-2020	115	4	we	-PRON-	PRP
zkie-zero-2020	115	5	derive	derive	VBP
zkie-zero-2020	115	6	a	a	DT
zkie-zero-2020	115	7	RDF	RDF	NNP
zkie-zero-2020	115	8	KB	KB	NNP
zkie-zero-2020	115	9	from	from	IN
zkie-zero-2020	115	10	their	-PRON-	PRP$
zkie-zero-2020	115	11	personography	personography	NN
zkie-zero-2020	115	12	,	,	,
zkie-zero-2020	115	13	for	for	IN
zkie-zero-2020	115	14	1641	1641	CD
zkie-zero-2020	115	15	we	-PRON-	PRP
zkie-zero-2020	115	16	derive	derive	VBP
zkie-zero-2020	115	17	a	a	DT
zkie-zero-2020	115	18	knowledge	knowledge	NN
zkie-zero-2020	115	19	base	base	NN
zkie-zero-2020	115	20	from	from	IN
zkie-zero-2020	115	21	the	the	DT
zkie-zero-2020	115	22	annotations	annotation	NNS
zkie-zero-2020	115	23	.	.	.
zkie-zero-2020	116	1	The	the	DT
zkie-zero-2020	116	2	exact	exact	JJ
zkie-zero-2020	116	3	processing	processing	NN
zkie-zero-2020	116	4	steps	step	NNS
zkie-zero-2020	116	5	as	as	RB
zkie-zero-2020	116	6	well	well	RB
zkie-zero-2020	116	7	as	as	IN
zkie-zero-2020	116	8	example	example	NN
zkie-zero-2020	116	9	texts	text	NNS
zkie-zero-2020	116	10	are	be	VBP
zkie-zero-2020	116	11	de-	de-	RB
zkie-zero-2020	116	12	scribed	scribe	VBN
zkie-zero-2020	116	13	in	in	IN
zkie-zero-2020	116	14	the	the	DT
zkie-zero-2020	116	15	appendix	appendix	NNP
zkie-zero-2020	116	16	.	.	.
zkie-zero-2020	117	1	The	the	DT
zkie-zero-2020	117	2	resulting	result	VBG
zkie-zero-2020	117	3	data	data	NN
zkie-zero-2020	117	4	sets	set	NNS
zkie-zero-2020	117	5	for	for	IN
zkie-zero-2020	117	6	WWO	wwo	NN
zkie-zero-2020	117	7	and	and	CC
zkie-zero-2020	117	8	1641	1641	CD
zkie-zero-2020	117	9	Depositions	Depositions	NNPS
zkie-zero-2020	117	10	are	be	VBP
zkie-zero-2020	117	11	also	also	RB
zkie-zero-2020	117	12	made	make	VBN
zkie-zero-2020	117	13	available	available	JJ
zkie-zero-2020	117	14	in	in	IN
zkie-zero-2020	117	15	the	the	DT
zkie-zero-2020	117	16	accompanying	accompany	VBG
zkie-zero-2020	117	17	code	code	NN
zkie-zero-2020	117	18	repository	repository	NN
zkie-zero-2020	117	19	.	.	.
zkie-zero-2020	118	1	AIDA	AIDA	NNP
zkie-zero-2020	118	2	-	-	HYPH
zkie-zero-2020	118	3	YAGO	YAGO	NNP
zkie-zero-2020	118	4	:	:	:
zkie-zero-2020	118	5	For	for	IN
zkie-zero-2020	118	6	validating	validate	VBG
zkie-zero-2020	118	7	our	-PRON-	PRP$
zkie-zero-2020	118	8	approach	approach	NN
zkie-zero-2020	118	9	,	,	,
zkie-zero-2020	118	10	we	-PRON-	PRP
zkie-zero-2020	118	11	evaluate	evaluate	VBP
zkie-zero-2020	118	12	on	on	IN
zkie-zero-2020	118	13	the	the	DT
zkie-zero-2020	118	14	AIDA	AIDA	NNP
zkie-zero-2020	118	15	-	-	HYPH
zkie-zero-2020	118	16	YAGO	YAGO	NNP
zkie-zero-2020	118	17	state	state	NN
zkie-zero-2020	118	18	-	-	HYPH
zkie-zero-2020	118	19	of	of	IN
zkie-zero-2020	118	20	-	-	HYPH
zkie-zero-2020	118	21	the	the	DT
zkie-zero-2020	118	22	art	art	NN
zkie-zero-2020	118	23	dataset	dataset	NN
zkie-zero-2020	118	24	introduced	introduce	VBN
zkie-zero-2020	118	25	by	by	IN
zkie-zero-2020	118	26	Hoffart	Hoffart	NNP
zkie-zero-2020	118	27	et	et	NNP
zkie-zero-2020	118	28	al	al	NNP
zkie-zero-2020	118	29	.	.	.
zkie-zero-2020	119	1	(	(	-LRB-
zkie-zero-2020	119	2	2011	2011	CD
zkie-zero-2020	119	3	)	)	-RRB-
zkie-zero-2020	119	4	.	.	.
zkie-zero-2020	120	1	Orig-	orig-	IN
zkie-zero-2020	120	2	inally	inally	RB
zkie-zero-2020	120	3	,	,	,
zkie-zero-2020	120	4	this	this	DT
zkie-zero-2020	120	5	dataset	dataset	NN
zkie-zero-2020	120	6	is	be	VBZ
zkie-zero-2020	120	7	linked	link	VBN
zkie-zero-2020	120	8	against	against	IN
zkie-zero-2020	120	9	YAGO	YAGO	NNP
zkie-zero-2020	120	10	and	and	CC
zkie-zero-2020	120	11	Wikipedia	Wikipedia	NNP
zkie-zero-2020	120	12	.	.	.
zkie-zero-2020	121	1	We	-PRON-	PRP
zkie-zero-2020	121	2	map	map	VBP
zkie-zero-2020	121	3	the	the	DT
zkie-zero-2020	121	4	Wikipedia	Wikipedia	NNP
zkie-zero-2020	121	5	URLs	url	NNS
zkie-zero-2020	121	6	to	to	IN
zkie-zero-2020	121	7	Wiki-	Wiki-	NNP
zkie-zero-2020	121	8	data	datum	NNS
zkie-zero-2020	121	9	and	and	CC
zkie-zero-2020	121	10	link	link	NN
zkie-zero-2020	121	11	against	against	IN
zkie-zero-2020	121	12	this	this	DT
zkie-zero-2020	121	13	KB	KB	NNP
zkie-zero-2020	121	14	,	,	,
zkie-zero-2020	121	15	as	as	IN
zkie-zero-2020	121	16	Wikidata	Wikidata	NNP
zkie-zero-2020	121	17	is	be	VBZ
zkie-zero-2020	121	18	avail-	avail-	JJ
zkie-zero-2020	121	19	able	able	JJ
zkie-zero-2020	121	20	in	in	IN
zkie-zero-2020	121	21	RDF	RDF	NNP
zkie-zero-2020	121	22	and	and	CC
zkie-zero-2020	121	23	the	the	DT
zkie-zero-2020	121	24	official	official	JJ
zkie-zero-2020	121	25	Wikidata	Wikidata	NNP
zkie-zero-2020	121	26	SPARQL	SPARQL	NNP
zkie-zero-2020	121	27	endpoint	endpoint	NN
zkie-zero-2020	121	28	offers	offer	VBZ
zkie-zero-2020	121	29	full	full	JJ
zkie-zero-2020	121	30	text	text	NN
zkie-zero-2020	121	31	search	search	NN
zkie-zero-2020	121	32	:	:	:
zkie-zero-2020	121	33	it	-PRON-	PRP
zkie-zero-2020	121	34	does	do	VBZ
zkie-zero-2020	121	35	not	not	RB
zkie-zero-2020	121	36	offer	offer	VB
zkie-zero-2020	121	37	fuzzy	fuzzy	JJ
zkie-zero-2020	121	38	search	search	NN
zkie-zero-2020	121	39	though	though	RB
zkie-zero-2020	121	40	.	.	.
zkie-zero-2020	122	1	Women	woman	NNS
zkie-zero-2020	122	2	Writers	Writers	NNPS
zkie-zero-2020	122	3	Online	online	RB
zkie-zero-2020	122	4	:	:	:
zkie-zero-2020	122	5	Women	Women	NNP
zkie-zero-2020	122	6	Writers	Writers	NNPS
zkie-zero-2020	122	7	On-	on-	CC
zkie-zero-2020	122	8	line3	line3	NN
zkie-zero-2020	122	9	is	be	VBZ
zkie-zero-2020	122	10	a	a	DT
zkie-zero-2020	122	11	collection	collection	NN
zkie-zero-2020	122	12	of	of	IN
zkie-zero-2020	122	13	texts	text	NNS
zkie-zero-2020	122	14	by	by	IN
zkie-zero-2020	122	15	pre	pre	JJ
zkie-zero-2020	122	16	-	-	JJ
zkie-zero-2020	122	17	Victorian	victorian	JJ
zkie-zero-2020	122	18	women	woman	NNS
zkie-zero-2020	122	19	writers	writer	NNS
zkie-zero-2020	122	20	.	.	.
zkie-zero-2020	123	1	It	-PRON-	PRP
zkie-zero-2020	123	2	includes	include	VBZ
zkie-zero-2020	123	3	texts	text	NNS
zkie-zero-2020	123	4	on	on	IN
zkie-zero-2020	123	5	a	a	DT
zkie-zero-2020	123	6	wide	wide	JJ
zkie-zero-2020	123	7	range	range	NN
zkie-zero-2020	123	8	of	of	IN
zkie-zero-2020	123	9	topics	topic	NNS
zkie-zero-2020	123	10	and	and	CC
zkie-zero-2020	123	11	from	from	IN
zkie-zero-2020	123	12	various	various	JJ
zkie-zero-2020	123	13	genres	genre	NNS
zkie-zero-2020	123	14	including	include	VBG
zkie-zero-2020	123	15	poems	poem	NNS
zkie-zero-2020	123	16	,	,	,
zkie-zero-2020	123	17	plays	play	NNS
zkie-zero-2020	123	18	,	,	,
zkie-zero-2020	123	19	and	and	CC
zkie-zero-2020	123	20	novels	novel	NNS
zkie-zero-2020	123	21	.	.	.
zkie-zero-2020	124	1	They	-PRON-	PRP
zkie-zero-2020	124	2	represent	represent	VBP
zkie-zero-2020	124	3	different	different	JJ
zkie-zero-2020	124	4	states	state	NNS
zkie-zero-2020	124	5	of	of	IN
zkie-zero-2020	124	6	the	the	DT
zkie-zero-2020	124	7	English	english	JJ
zkie-zero-2020	124	8	language	language	NN
zkie-zero-2020	124	9	between	between	IN
zkie-zero-2020	124	10	1400	1400	CD
zkie-zero-2020	124	11	and	and	CC
zkie-zero-2020	124	12	1850	1850	CD
zkie-zero-2020	124	13	.	.	.
zkie-zero-2020	125	1	A	a	DT
zkie-zero-2020	125	2	subset	subset	NN
zkie-zero-2020	125	3	of	of	IN
zkie-zero-2020	125	4	documents	document	NNS
zkie-zero-2020	125	5	has	have	VBZ
zkie-zero-2020	125	6	been	be	VBN
zkie-zero-2020	125	7	annotated	annotate	VBN
zkie-zero-2020	125	8	with	with	IN
zkie-zero-2020	125	9	named	name	VBN
zkie-zero-2020	125	10	entities	entity	NNS
zkie-zero-2020	125	11	(	(	-LRB-
zkie-zero-2020	125	12	persons	person	NNS
zkie-zero-2020	125	13	,	,	,
zkie-zero-2020	125	14	works	work	NNS
zkie-zero-2020	125	15	,	,	,
zkie-zero-2020	125	16	places	place	NNS
zkie-zero-2020	125	17	)	)	-RRB-
zkie-zero-2020	125	18	(	(	-LRB-
zkie-zero-2020	125	19	Melson	Melson	NNP
zkie-zero-2020	125	20	and	and	CC
zkie-zero-2020	125	21	Flanders	Flanders	NNP
zkie-zero-2020	125	22	,	,	,
zkie-zero-2020	125	23	2010	2010	CD
zkie-zero-2020	125	24	)	)	-RRB-
zkie-zero-2020	125	25	.	.	.
zkie-zero-2020	126	1	Persons	person	NNS
zkie-zero-2020	126	2	have	have	VBP
zkie-zero-2020	126	3	also	also	RB
zkie-zero-2020	126	4	been	be	VBN
zkie-zero-2020	126	5	linked	link	VBN
zkie-zero-2020	126	6	to	to	TO
zkie-zero-2020	126	7	create	create	VB
zkie-zero-2020	126	8	a	a	DT
zkie-zero-2020	126	9	personography	personography	NN
zkie-zero-2020	126	10	,	,	,
zkie-zero-2020	126	11	a	a	DT
zkie-zero-2020	126	12	structured	structure	VBN
zkie-zero-2020	126	13	represen-	represen-	JJ
zkie-zero-2020	126	14	tation	tation	NN
zkie-zero-2020	126	15	of	of	IN
zkie-zero-2020	126	16	persons	person	NNS
zkie-zero-2020	126	17	’	’	POS
zkie-zero-2020	126	18	biographies	biography	NNS
zkie-zero-2020	126	19	containing	contain	VBG
zkie-zero-2020	126	20	names	name	NNS
zkie-zero-2020	126	21	,	,	,
zkie-zero-2020	126	22	titles	title	NNS
zkie-zero-2020	126	23	,	,	,
zkie-zero-2020	126	24	time	time	NN
zkie-zero-2020	126	25	and	and	CC
zkie-zero-2020	126	26	place	place	NN
zkie-zero-2020	126	27	of	of	IN
zkie-zero-2020	126	28	birth	birth	NN
zkie-zero-2020	126	29	and	and	CC
zkie-zero-2020	126	30	death	death	NN
zkie-zero-2020	126	31	.	.	.
zkie-zero-2020	127	1	The	the	DT
zkie-zero-2020	127	2	texts	text	NNS
zkie-zero-2020	127	3	are	be	VBP
zkie-zero-2020	127	4	challenging	challenge	VBG
zkie-zero-2020	127	5	to	to	TO
zkie-zero-2020	127	6	disambiguate	disambiguate	VB
zkie-zero-2020	127	7	due	due	IN
zkie-zero-2020	127	8	to	to	IN
zkie-zero-2020	127	9	spelling	spelling	NN
zkie-zero-2020	127	10	variance	variance	NN
zkie-zero-2020	127	11	,	,	,
zkie-zero-2020	127	12	ciphering	cipher	VBG
zkie-zero-2020	127	13	of	of	IN
zkie-zero-2020	127	14	names	name	NNS
zkie-zero-2020	127	15	and	and	CC
zkie-zero-2020	127	16	a	a	DT
zkie-zero-2020	127	17	lack	lack	NN
zkie-zero-2020	127	18	of	of	IN
zkie-zero-2020	127	19	stan-	stan-	JJ
zkie-zero-2020	127	20	dardized	dardize	VBN
zkie-zero-2020	127	21	orthography	orthography	NN
zkie-zero-2020	127	22	.	.	.
zkie-zero-2020	128	1	Sometimes	sometimes	RB
zkie-zero-2020	128	2	,	,	,
zkie-zero-2020	128	3	people	people	NNS
zkie-zero-2020	128	4	are	be	VBP
zkie-zero-2020	128	5	not	not	RB
zkie-zero-2020	128	6	referred	refer	VBN
zkie-zero-2020	128	7	to	to	IN
zkie-zero-2020	128	8	by	by	IN
zkie-zero-2020	128	9	name	name	NN
zkie-zero-2020	128	10	but	but	CC
zkie-zero-2020	128	11	by	by	IN
zkie-zero-2020	128	12	rank	rank	NN
zkie-zero-2020	128	13	or	or	CC
zkie-zero-2020	128	14	function	function	NN
zkie-zero-2020	128	15	,	,	,
zkie-zero-2020	128	16	e.g.	e.g.	RB
zkie-zero-2020	129	1	the	the	DT
zkie-zero-2020	129	2	king	king	NN
zkie-zero-2020	129	3	.	.	.
zkie-zero-2020	130	1	This	this	DT
zkie-zero-2020	130	2	dataset	dataset	NN
zkie-zero-2020	130	3	is	be	VBZ
zkie-zero-2020	130	4	interesting	interesting	JJ
zkie-zero-2020	130	5	,	,	,
zkie-zero-2020	130	6	as	as	IN
zkie-zero-2020	130	7	it	-PRON-	PRP
zkie-zero-2020	130	8	contains	contain	VBZ
zkie-zero-2020	130	9	doc-	doc-	NN
zkie-zero-2020	130	10	uments	ument	NNS
zkie-zero-2020	130	11	with	with	IN
zkie-zero-2020	130	12	heterogeneous	heterogeneous	JJ
zkie-zero-2020	130	13	topics	topic	NNS
zkie-zero-2020	130	14	and	and	CC
zkie-zero-2020	130	15	text	text	NN
zkie-zero-2020	130	16	genres	genre	NNS
zkie-zero-2020	130	17	,	,	,
zkie-zero-2020	130	18	causing	cause	VBG
zkie-zero-2020	130	19	low	low	JJ
zkie-zero-2020	130	20	redundancy	redundancy	NN
zkie-zero-2020	130	21	.	.	.
zkie-zero-2020	131	1	1641	1641	CD
zkie-zero-2020	131	2	Depositions	Depositions	NNPS
zkie-zero-2020	131	3	:	:	:
zkie-zero-2020	131	4	The	the	DT
zkie-zero-2020	131	5	1641	1641	CD
zkie-zero-2020	131	6	Depositions4	Depositions4	NNP
zkie-zero-2020	131	7	con-	con-	NN
zkie-zero-2020	131	8	tain	tain	VBP
zkie-zero-2020	131	9	legal	legal	JJ
zkie-zero-2020	131	10	texts	text	NNS
zkie-zero-2020	131	11	in	in	IN
zkie-zero-2020	131	12	form	form	NN
zkie-zero-2020	131	13	of	of	IN
zkie-zero-2020	131	14	court	court	NN
zkie-zero-2020	131	15	witness	witness	NN
zkie-zero-2020	131	16	statements	statement	NNS
zkie-zero-2020	131	17	recorded	record	VBN
zkie-zero-2020	131	18	after	after	IN
zkie-zero-2020	131	19	the	the	DT
zkie-zero-2020	131	20	Irish	Irish	NNP
zkie-zero-2020	131	21	Rebellion	Rebellion	NNP
zkie-zero-2020	131	22	of	of	IN
zkie-zero-2020	131	23	1641	1641	CD
zkie-zero-2020	131	24	.	.	.
zkie-zero-2020	132	1	In	in	IN
zkie-zero-2020	132	2	this	this	DT
zkie-zero-2020	132	3	conflict	conflict	NN
zkie-zero-2020	132	4	,	,	,
zkie-zero-2020	132	5	Irish	Irish	NNP
zkie-zero-2020	132	6	and	and	CC
zkie-zero-2020	132	7	English	English	NNP
zkie-zero-2020	132	8	Catholics	Catholics	NNPS
zkie-zero-2020	132	9	revolted	revolt	VBD
zkie-zero-2020	132	10	against	against	IN
zkie-zero-2020	132	11	English	English	NNP
zkie-zero-2020	132	12	and	and	CC
zkie-zero-2020	132	13	Scottish	Scottish	NNP
zkie-zero-2020	132	14	Protestants	Protestants	NNPS
zkie-zero-2020	132	15	and	and	CC
zkie-zero-2020	132	16	their	-PRON-	PRP$
zkie-zero-2020	132	17	colonization	colonization	NN
zkie-zero-2020	132	18	of	of	IN
zkie-zero-2020	132	19	Ireland	Ireland	NNP
zkie-zero-2020	132	20	.	.	.
zkie-zero-2020	133	1	It	-PRON-	PRP
zkie-zero-2020	133	2	lasted	last	VBD
zkie-zero-2020	133	3	over	over	IN
zkie-zero-2020	133	4	10	10	CD
zkie-zero-2020	133	5	years	year	NNS
zkie-zero-2020	133	6	and	and	CC
zkie-zero-2020	133	7	ended	end	VBD
zkie-zero-2020	133	8	with	with	IN
zkie-zero-2020	133	9	the	the	DT
zkie-zero-2020	133	10	Irish	Irish	NNP
zkie-zero-2020	133	11	Catholics	Catholics	NNPS
zkie-zero-2020	133	12	’	’	POS
zkie-zero-2020	133	13	defeat	defeat	NN
zkie-zero-2020	133	14	and	and	CC
zkie-zero-2020	133	15	the	the	DT
zkie-zero-2020	133	16	for-	for-	NN
zkie-zero-2020	133	17	eign	eign	JJ
zkie-zero-2020	133	18	rule	rule	NN
zkie-zero-2020	133	19	of	of	IN
zkie-zero-2020	133	20	Ireland	Ireland	NNP
zkie-zero-2020	133	21	.	.	.
zkie-zero-2020	134	1	The	the	DT
zkie-zero-2020	134	2	depositions	deposition	NNS
zkie-zero-2020	134	3	have	have	VBP
zkie-zero-2020	134	4	been	be	VBN
zkie-zero-2020	134	5	transcribed	transcribe	VBN
zkie-zero-2020	134	6	from	from	IN
zkie-zero-2020	134	7	17th	17th	JJ
zkie-zero-2020	134	8	century	century	NN
zkie-zero-2020	134	9	handwriting	handwriting	NN
zkie-zero-2020	134	10	,	,	,
zkie-zero-2020	134	11	keep-	keep-	NNP
zkie-zero-2020	134	12	ing	e	VBG
zkie-zero-2020	134	13	the	the	DT
zkie-zero-2020	134	14	old	old	JJ
zkie-zero-2020	134	15	language	language	NN
zkie-zero-2020	134	16	and	and	CC
zkie-zero-2020	134	17	orthography	orthography	NN
zkie-zero-2020	134	18	.	.	.
zkie-zero-2020	135	1	These	these	DT
zkie-zero-2020	135	2	doc-	doc-	NN
zkie-zero-2020	135	3	uments	ument	NNS
zkie-zero-2020	135	4	have	have	VBP
zkie-zero-2020	135	5	been	be	VBN
zkie-zero-2020	135	6	used	use	VBN
zkie-zero-2020	135	7	to	to	TO
zkie-zero-2020	135	8	analyze	analyze	VB
zkie-zero-2020	135	9	the	the	DT
zkie-zero-2020	135	10	rebellion	rebellion	NN
zkie-zero-2020	135	11	,	,	,
zkie-zero-2020	135	12	perform	perform	VB
zkie-zero-2020	135	13	cold	cold	JJ
zkie-zero-2020	135	14	case	case	NN
zkie-zero-2020	135	15	reviews	review	NNS
zkie-zero-2020	135	16	of	of	IN
zkie-zero-2020	135	17	the	the	DT
zkie-zero-2020	135	18	atrocities	atrocity	NNS
zkie-zero-2020	135	19	commit-	commit-	NNP
zkie-zero-2020	135	20	ted	ted	NNP
zkie-zero-2020	135	21	and	and	CC
zkie-zero-2020	135	22	to	to	TO
zkie-zero-2020	135	23	gain	gain	VB
zkie-zero-2020	135	24	insights	insight	NNS
zkie-zero-2020	135	25	into	into	IN
zkie-zero-2020	135	26	contemporary	contemporary	JJ
zkie-zero-2020	135	27	life	life	NN
zkie-zero-2020	135	28	of	of	IN
zkie-zero-2020	135	29	this	this	DT
zkie-zero-2020	135	30	era	era	NN
zkie-zero-2020	135	31	.	.	.
zkie-zero-2020	136	1	Part	part	NN
zkie-zero-2020	136	2	of	of	IN
zkie-zero-2020	136	3	the	the	DT
zkie-zero-2020	136	4	documents	document	NNS
zkie-zero-2020	136	5	have	have	VBP
zkie-zero-2020	136	6	been	be	VBN
zkie-zero-2020	136	7	annotated	annotate	VBN
zkie-zero-2020	136	8	3	3	CD
zkie-zero-2020	136	9	https://www.wwp.northeastern.edu/wwo	https://www.wwp.northeastern.edu/wwo	NNP
zkie-zero-2020	136	10	4	4	CD
zkie-zero-2020	136	11	http://1641.tcd.ie/	http://1641.tcd.ie/	CD
zkie-zero-2020	136	12	https://www.wwp.northeastern.edu/wwo	https://www.wwp.northeastern.edu/wwo	NN
zkie-zero-2020	136	13	http://1641.tcd.ie/	http://1641.tcd.ie/	ADD
zkie-zero-2020	136	14	6986	6986	CD
zkie-zero-2020	136	15	Table	table	NN
zkie-zero-2020	136	16	2	2	CD
zkie-zero-2020	136	17	:	:	:
zkie-zero-2020	136	18	Data	data	NN
zkie-zero-2020	136	19	statistics	statistic	NNS
zkie-zero-2020	136	20	of	of	IN
zkie-zero-2020	136	21	the	the	DT
zkie-zero-2020	136	22	three	three	CD
zkie-zero-2020	136	23	used	use	VBN
zkie-zero-2020	136	24	datasets	dataset	NNS
zkie-zero-2020	136	25	:	:	:
zkie-zero-2020	136	26	Total	total	JJ
zkie-zero-2020	136	27	number	number	NN
zkie-zero-2020	136	28	of	of	IN
zkie-zero-2020	136	29	Documents	document	NNS
zkie-zero-2020	136	30	,	,	,
zkie-zero-2020	136	31	Tokens	Tokens	NNPS
zkie-zero-2020	136	32	,	,	,
zkie-zero-2020	136	33	Entities	entity	NNS
zkie-zero-2020	136	34	,	,	,
zkie-zero-2020	136	35	average	average	JJ
zkie-zero-2020	136	36	number	number	NN
zkie-zero-2020	136	37	of	of	IN
zkie-zero-2020	136	38	Entities	Entities	NNPS
zkie-zero-2020	136	39	per	per	IN
zkie-zero-2020	136	40	Sentence	Sentence	NNP
zkie-zero-2020	136	41	,	,	,
zkie-zero-2020	136	42	%	%	NN
zkie-zero-2020	136	43	of	of	IN
zkie-zero-2020	136	44	entities	entity	NNS
zkie-zero-2020	136	45	that	that	WDT
zkie-zero-2020	136	46	are	be	VBP
zkie-zero-2020	136	47	not	not	RB
zkie-zero-2020	136	48	linked	link	VBN
zkie-zero-2020	136	49	.	.	.
zkie-zero-2020	137	1	We	-PRON-	PRP
zkie-zero-2020	137	2	also	also	RB
zkie-zero-2020	137	3	report	report	VBP
zkie-zero-2020	137	4	the	the	DT
zkie-zero-2020	137	5	average	average	JJ
zkie-zero-2020	137	6	number	number	NN
zkie-zero-2020	137	7	of	of	IN
zkie-zero-2020	137	8	entities	entity	NNS
zkie-zero-2020	137	9	linked	link	VBN
zkie-zero-2020	137	10	to	to	IN
zkie-zero-2020	137	11	a	a	DT
zkie-zero-2020	137	12	mention	mention	NN
zkie-zero-2020	137	13	,	,	,
zkie-zero-2020	137	14	the	the	DT
zkie-zero-2020	137	15	average	average	JJ
zkie-zero-2020	137	16	number	number	NN
zkie-zero-2020	137	17	of	of	IN
zkie-zero-2020	137	18	candidates	candidate	NNS
zkie-zero-2020	137	19	when	when	WRB
zkie-zero-2020	137	20	searching	search	VBG
zkie-zero-2020	137	21	for	for	IN
zkie-zero-2020	137	22	a	a	DT
zkie-zero-2020	137	23	mention	mention	NN
zkie-zero-2020	137	24	in	in	IN
zkie-zero-2020	137	25	the	the	DT
zkie-zero-2020	137	26	KB	KB	NNP
zkie-zero-2020	137	27	and	and	CC
zkie-zero-2020	137	28	the	the	DT
zkie-zero-2020	137	29	Gini	Gini	NNP
zkie-zero-2020	137	30	coefficient	coefficient	NN
zkie-zero-2020	137	31	which	which	WDT
zkie-zero-2020	137	32	measures	measure	NNS
zkie-zero-2020	137	33	how	how	WRB
zkie-zero-2020	137	34	balanced	balanced	JJ
zkie-zero-2020	137	35	the	the	DT
zkie-zero-2020	137	36	entity	entity	NN
zkie-zero-2020	137	37	distribution	distribution	NN
zkie-zero-2020	137	38	is	be	VBZ
zkie-zero-2020	137	39	.	.	.
zkie-zero-2020	138	1	Corpus	Corpus	NNP
zkie-zero-2020	138	2	#	#	NN
zkie-zero-2020	138	3	D	d	NN
zkie-zero-2020	138	4	#	#	$
zkie-zero-2020	138	5	T	t	NN
zkie-zero-2020	138	6	#	#	$
zkie-zero-2020	138	7	E	e	NN
zkie-zero-2020	138	8	#	#	NN
zkie-zero-2020	138	9	E	e	NN
zkie-zero-2020	138	10	/	/	SYM
zkie-zero-2020	138	11	S	S	NNP
zkie-zero-2020	138	12	%	%	NN
zkie-zero-2020	138	13	NIL	NIL	NNP
zkie-zero-2020	138	14	Avg	Avg	NNP
zkie-zero-2020	138	15	.	.	.
zkie-zero-2020	139	1	Amb	Amb	NNP
zkie-zero-2020	139	2	.	.	.
zkie-zero-2020	140	1	Avg	Avg	NNP
zkie-zero-2020	140	2	.	.	.
zkie-zero-2020	141	1	#	#	NNP
zkie-zero-2020	141	2	Cand	Cand	NNP
zkie-zero-2020	141	3	.	.	.
zkie-zero-2020	142	1	Gini	Gini	NNP
zkie-zero-2020	142	2	AIDA	AIDA	NNP
zkie-zero-2020	142	3	1393	1393	CD
zkie-zero-2020	142	4	301,418	301,418	CD
zkie-zero-2020	142	5	34,929	34,929	CD
zkie-zero-2020	142	6	1.59	1.59	CD
zkie-zero-2020	142	7	20.37	20.37	CD
zkie-zero-2020	142	8	1.08	1.08	CD
zkie-zero-2020	142	9	6.98	6.98	CD
zkie-zero-2020	142	10	0.73	0.73	CD
zkie-zero-2020	142	11	WWO	WWO	NNP
zkie-zero-2020	142	12	74	74	CD
zkie-zero-2020	142	13	1,461,401	1,461,401	CD
zkie-zero-2020	142	14	14,651	14,651	CD
zkie-zero-2020	142	15	0.34	0.34	CD
zkie-zero-2020	142	16	7.42	7.42	CD
zkie-zero-2020	142	17	1.08	1.08	CD
zkie-zero-2020	142	18	16.66	16.66	CD
zkie-zero-2020	142	19	0.56	0.56	CD
zkie-zero-2020	142	20	1641	1641	CD
zkie-zero-2020	142	21	16	16	CD
zkie-zero-2020	142	22	11,895	11,895	CD
zkie-zero-2020	142	23	480	480	CD
zkie-zero-2020	142	24	2.40	2.40	CD
zkie-zero-2020	142	25	0.0	0.0	CD
zkie-zero-2020	142	26	1.01	1.01	CD
zkie-zero-2020	142	27	36.29	36.29	CD
zkie-zero-2020	142	28	0.44	0.44	CD
zkie-zero-2020	142	29	with	with	IN
zkie-zero-2020	142	30	named	name	VBN
zkie-zero-2020	142	31	entities	entity	NNS
zkie-zero-2020	142	32	that	that	WDT
zkie-zero-2020	142	33	are	be	VBP
zkie-zero-2020	142	34	linked	link	VBN
zkie-zero-2020	142	35	to	to	IN
zkie-zero-2020	142	36	DBPedia	DBPedia	NNP
zkie-zero-2020	142	37	(	(	-LRB-
zkie-zero-2020	142	38	Munnelly	munnelly	RB
zkie-zero-2020	142	39	and	and	CC
zkie-zero-2020	142	40	Lawless	Lawless	NNP
zkie-zero-2020	142	41	,	,	,
zkie-zero-2020	142	42	2018b	2018b	CD
zkie-zero-2020	142	43	)	)	-RRB-
zkie-zero-2020	142	44	.	.	.
zkie-zero-2020	143	1	As	as	IN
zkie-zero-2020	143	2	the	the	DT
zkie-zero-2020	143	3	coverage	coverage	NN
zkie-zero-2020	143	4	of	of	IN
zkie-zero-2020	143	5	DBPedia	DBPedia	NNP
zkie-zero-2020	143	6	was	be	VBD
zkie-zero-2020	143	7	not	not	RB
zkie-zero-2020	143	8	sufficient	sufficient	JJ
zkie-zero-2020	143	9	(	(	-LRB-
zkie-zero-2020	143	10	only	only	RB
zkie-zero-2020	143	11	around	around	IN
zkie-zero-2020	143	12	20	20	CD
zkie-zero-2020	143	13	%	%	NN
zkie-zero-2020	143	14	of	of	IN
zkie-zero-2020	143	15	the	the	DT
zkie-zero-2020	143	16	entities	entity	NNS
zkie-zero-2020	143	17	are	be	VBP
zkie-zero-2020	143	18	in	in	IN
zkie-zero-2020	143	19	DBPedia	DBPedia	NNP
zkie-zero-2020	143	20	)	)	-RRB-
zkie-zero-2020	143	21	,	,	,
zkie-zero-2020	143	22	we	-PRON-	PRP
zkie-zero-2020	143	23	manually	manually	RB
zkie-zero-2020	143	24	cre-	cre-	VBP
zkie-zero-2020	143	25	ated	ate	VBD
zkie-zero-2020	143	26	a	a	DT
zkie-zero-2020	143	27	domain	domain	NN
zkie-zero-2020	143	28	specific	specific	JJ
zkie-zero-2020	143	29	knowledge	knowledge	NN
zkie-zero-2020	143	30	base	base	NN
zkie-zero-2020	143	31	for	for	IN
zkie-zero-2020	143	32	this	this	DT
zkie-zero-2020	143	33	data	data	NN
zkie-zero-2020	143	34	set	set	NN
zkie-zero-2020	143	35	containing	contain	VBG
zkie-zero-2020	143	36	places	place	NNS
zkie-zero-2020	143	37	and	and	CC
zkie-zero-2020	143	38	people	people	NNS
zkie-zero-2020	143	39	mentioned	mention	VBD
zkie-zero-2020	143	40	.	.	.
zkie-zero-2020	144	1	To	to	TO
zkie-zero-2020	144	2	increase	increase	VB
zkie-zero-2020	144	3	difficulty	difficulty	NN
zkie-zero-2020	144	4	and	and	CC
zkie-zero-2020	144	5	reduce	reduce	VB
zkie-zero-2020	144	6	overfitting	overfitting	NN
zkie-zero-2020	144	7	,	,	,
zkie-zero-2020	144	8	we	-PRON-	PRP
zkie-zero-2020	144	9	added	add	VBD
zkie-zero-2020	144	10	additional	additional	JJ
zkie-zero-2020	144	11	related	relate	VBN
zkie-zero-2020	144	12	entities	entity	NNS
zkie-zero-2020	144	13	from	from	IN
zkie-zero-2020	144	14	DBPedia	DBPedia	NNP
zkie-zero-2020	144	15	.	.	.
zkie-zero-2020	145	1	The	the	DT
zkie-zero-2020	145	2	num-	num-	JJ
zkie-zero-2020	145	3	ber	ber	NN
zkie-zero-2020	145	4	of	of	IN
zkie-zero-2020	145	5	persons	person	NNS
zkie-zero-2020	145	6	increases	increase	VBZ
zkie-zero-2020	145	7	thereby	thereby	RB
zkie-zero-2020	145	8	by	by	IN
zkie-zero-2020	145	9	tenfold	tenfold	NN
zkie-zero-2020	145	10	(	(	-LRB-
zkie-zero-2020	145	11	130	130	CD
zkie-zero-2020	145	12	→	→	SYM
zkie-zero-2020	145	13	1383	1383	CD
zkie-zero-2020	145	14	)	)	-RRB-
zkie-zero-2020	145	15	and	and	CC
zkie-zero-2020	145	16	the	the	DT
zkie-zero-2020	145	17	number	number	NN
zkie-zero-2020	145	18	of	of	IN
zkie-zero-2020	145	19	places	place	NNS
zkie-zero-2020	145	20	by	by	IN
zkie-zero-2020	145	21	twentyfold	twentyfold	NN
zkie-zero-2020	145	22	(	(	-LRB-
zkie-zero-2020	145	23	99	99	CD
zkie-zero-2020	145	24	→	→	SYM
zkie-zero-2020	145	25	2119	2119	CD
zkie-zero-2020	145	26	)	)	-RRB-
zkie-zero-2020	145	27	.	.	.
zkie-zero-2020	146	1	Details	detail	NNS
zkie-zero-2020	146	2	for	for	IN
zkie-zero-2020	146	3	that	that	DT
zkie-zero-2020	146	4	can	can	MD
zkie-zero-2020	146	5	be	be	VB
zkie-zero-2020	146	6	found	find	VBN
zkie-zero-2020	146	7	in	in	IN
zkie-zero-2020	146	8	Ap-	Ap-	NNP
zkie-zero-2020	146	9	pendix	pendix	NNP
zkie-zero-2020	146	10	A.1	A.1	NNP
zkie-zero-2020	146	11	.	.	.
zkie-zero-2020	147	1	While	while	IN
zkie-zero-2020	147	2	generating	generate	VBG
zkie-zero-2020	147	3	a	a	DT
zkie-zero-2020	147	4	KB	kb	NN
zkie-zero-2020	147	5	from	from	IN
zkie-zero-2020	147	6	gold	gold	NN
zkie-zero-2020	147	7	data	datum	NNS
zkie-zero-2020	147	8	is	be	VBZ
zkie-zero-2020	147	9	not	not	RB
zkie-zero-2020	147	10	ideal	ideal	JJ
zkie-zero-2020	147	11	,	,	,
zkie-zero-2020	147	12	creating	create	VBG
zkie-zero-2020	147	13	or	or	CC
zkie-zero-2020	147	14	completing	complete	VBG
zkie-zero-2020	147	15	a	a	DT
zkie-zero-2020	147	16	knowledge	knowledge	NN
zkie-zero-2020	147	17	base	base	NN
zkie-zero-2020	147	18	during	during	IN
zkie-zero-2020	147	19	annotation	annotation	NN
zkie-zero-2020	147	20	is	be	VBZ
zkie-zero-2020	147	21	not	not	RB
zkie-zero-2020	147	22	uncommon	uncommon	JJ
zkie-zero-2020	147	23	(	(	-LRB-
zkie-zero-2020	147	24	see	see	VB
zkie-zero-2020	147	25	e.g.	e.g.	RB
zkie-zero-2020	148	1	Wolfe	Wolfe	NNP
zkie-zero-2020	148	2	et	et	FW
zkie-zero-2020	148	3	al	al	NNP
zkie-zero-2020	148	4	.	.	NNP
zkie-zero-2020	148	5	,	,	,
zkie-zero-2020	148	6	2015	2015	CD
zkie-zero-2020	148	7	)	)	-RRB-
zkie-zero-2020	148	8	.	.	.
zkie-zero-2020	149	1	The	the	DT
zkie-zero-2020	149	2	texts	text	NNS
zkie-zero-2020	149	3	are	be	VBP
zkie-zero-2020	149	4	difficult	difficult	JJ
zkie-zero-2020	149	5	to	to	TO
zkie-zero-2020	149	6	disam-	disam-	VB
zkie-zero-2020	149	7	biguate	biguate	NN
zkie-zero-2020	149	8	due	due	IN
zkie-zero-2020	149	9	to	to	IN
zkie-zero-2020	149	10	the	the	DT
zkie-zero-2020	149	11	same	same	JJ
zkie-zero-2020	149	12	reasons	reason	NNS
zkie-zero-2020	149	13	as	as	IN
zkie-zero-2020	149	14	for	for	IN
zkie-zero-2020	149	15	WWO	wwo	NN
zkie-zero-2020	149	16	.	.	.
zkie-zero-2020	150	1	The	the	DT
zkie-zero-2020	150	2	depositions	deposition	NNS
zkie-zero-2020	150	3	are	be	VBP
zkie-zero-2020	150	4	interesting	interesting	JJ
zkie-zero-2020	150	5	,	,	,
zkie-zero-2020	150	6	as	as	IN
zkie-zero-2020	150	7	they	-PRON-	PRP
zkie-zero-2020	150	8	contain	contain	VBP
zkie-zero-2020	150	9	docu-	docu-	NN
zkie-zero-2020	150	10	ments	ment	NNS
zkie-zero-2020	150	11	from	from	IN
zkie-zero-2020	150	12	the	the	DT
zkie-zero-2020	150	13	same	same	JJ
zkie-zero-2020	150	14	domain	domain	NN
zkie-zero-2020	150	15	(	(	-LRB-
zkie-zero-2020	150	16	witness	witness	NN
zkie-zero-2020	150	17	reports	report	NNS
zkie-zero-2020	150	18	)	)	-RRB-
zkie-zero-2020	150	19	,	,	,
zkie-zero-2020	150	20	but	but	CC
zkie-zero-2020	150	21	feature	feature	VBP
zkie-zero-2020	150	22	many	many	JJ
zkie-zero-2020	150	23	different	different	JJ
zkie-zero-2020	150	24	actors	actor	NNS
zkie-zero-2020	150	25	and	and	CC
zkie-zero-2020	150	26	events	event	NNS
zkie-zero-2020	150	27	.	.	.
zkie-zero-2020	151	1	Table	table	NN
zkie-zero-2020	151	2	2	2	CD
zkie-zero-2020	151	3	contains	contain	VBZ
zkie-zero-2020	151	4	several	several	JJ
zkie-zero-2020	151	5	statistics	statistic	NNS
zkie-zero-2020	151	6	regarding	regard	VBG
zkie-zero-2020	151	7	the	the	DT
zkie-zero-2020	151	8	three	three	CD
zkie-zero-2020	151	9	datasets	dataset	NNS
zkie-zero-2020	151	10	.	.	.
zkie-zero-2020	152	1	AIDA	AIDA	NNP
zkie-zero-2020	152	2	and	and	CC
zkie-zero-2020	152	3	1641	1641	CD
zkie-zero-2020	152	4	contain	contain	NN
zkie-zero-2020	152	5	on	on	IN
zkie-zero-2020	152	6	aver-	aver-	JJ
zkie-zero-2020	152	7	age	age	NN
zkie-zero-2020	152	8	at	at	RB
zkie-zero-2020	152	9	least	least	RBS
zkie-zero-2020	152	10	one	one	CD
zkie-zero-2020	152	11	entity	entity	NN
zkie-zero-2020	152	12	per	per	IN
zkie-zero-2020	152	13	sentence	sentence	NN
zkie-zero-2020	152	14	,	,	,
zkie-zero-2020	152	15	whereas	whereas	IN
zkie-zero-2020	152	16	WWO	wwo	NN
zkie-zero-2020	152	17	,	,	,
zkie-zero-2020	152	18	while	while	IN
zkie-zero-2020	152	19	larger	large	JJR
zkie-zero-2020	152	20	,	,	,
zkie-zero-2020	152	21	is	be	VBZ
zkie-zero-2020	152	22	only	only	RB
zkie-zero-2020	152	23	sparsely	sparsely	RB
zkie-zero-2020	152	24	annotated	annotate	VBN
zkie-zero-2020	152	25	.	.	.
zkie-zero-2020	153	1	In	in	IN
zkie-zero-2020	153	2	con-	con-	NN
zkie-zero-2020	153	3	trast	trast	NN
zkie-zero-2020	153	4	to	to	IN
zkie-zero-2020	153	5	the	the	DT
zkie-zero-2020	153	6	other	other	JJ
zkie-zero-2020	153	7	two	two	CD
zkie-zero-2020	153	8	,	,	,
zkie-zero-2020	153	9	1641	1641	CD
zkie-zero-2020	153	10	contains	contain	VBZ
zkie-zero-2020	153	11	no	no	DT
zkie-zero-2020	153	12	entities	entity	NNS
zkie-zero-2020	153	13	linked	link	VBN
zkie-zero-2020	153	14	to	to	IN
zkie-zero-2020	153	15	NIL	NIL	NNP
zkie-zero-2020	153	16	.	.	.
zkie-zero-2020	154	1	This	this	DT
zkie-zero-2020	154	2	is	be	VBZ
zkie-zero-2020	154	3	caused	cause	VBN
zkie-zero-2020	154	4	by	by	IN
zkie-zero-2020	154	5	the	the	DT
zkie-zero-2020	154	6	fact	fact	NN
zkie-zero-2020	154	7	that	that	IN
zkie-zero-2020	154	8	we	-PRON-	PRP
zkie-zero-2020	154	9	created	create	VBD
zkie-zero-2020	154	10	the	the	DT
zkie-zero-2020	154	11	KB	KB	NNP
zkie-zero-2020	154	12	for	for	IN
zkie-zero-2020	154	13	1641	1641	CD
zkie-zero-2020	154	14	from	from	IN
zkie-zero-2020	154	15	the	the	DT
zkie-zero-2020	154	16	gold	gold	NN
zkie-zero-2020	154	17	annota-	annota-	JJ
zkie-zero-2020	154	18	tions	tion	NNS
zkie-zero-2020	154	19	and	and	CC
zkie-zero-2020	154	20	for	for	IN
zkie-zero-2020	154	21	entities	entity	NNS
zkie-zero-2020	154	22	previously	previously	RB
zkie-zero-2020	154	23	NIL	NIL	NNP
zkie-zero-2020	154	24	,	,	,
zkie-zero-2020	154	25	new	new	JJ
zkie-zero-2020	154	26	entities	entity	NNS
zkie-zero-2020	154	27	were	be	VBD
zkie-zero-2020	154	28	created	create	VBN
zkie-zero-2020	154	29	by	by	IN
zkie-zero-2020	154	30	hand	hand	NN
zkie-zero-2020	154	31	;	;	:
zkie-zero-2020	154	32	before	before	IN
zkie-zero-2020	154	33	that	that	DT
zkie-zero-2020	154	34	,	,	,
zkie-zero-2020	154	35	the	the	DT
zkie-zero-2020	154	36	original	original	JJ
zkie-zero-2020	154	37	corpus	corpus	NN
zkie-zero-2020	154	38	linking	link	VBG
zkie-zero-2020	154	39	to	to	IN
zkie-zero-2020	154	40	DBPedia	DBPedia	NNP
zkie-zero-2020	154	41	had	have	VBD
zkie-zero-2020	154	42	77	77	CD
zkie-zero-2020	154	43	%	%	NN
zkie-zero-2020	154	44	NIL	NIL	NNP
zkie-zero-2020	154	45	annota-	annota-	JJ
zkie-zero-2020	154	46	tions	tion	NNS
zkie-zero-2020	154	47	.	.	.
zkie-zero-2020	155	1	The	the	DT
zkie-zero-2020	155	2	average	average	JJ
zkie-zero-2020	155	3	ambiguity	ambiguity	NN
zkie-zero-2020	155	4	,	,	,
zkie-zero-2020	155	5	that	that	RB
zkie-zero-2020	155	6	is	is	RB
zkie-zero-2020	155	7	,	,	,
zkie-zero-2020	155	8	how	how	WRB
zkie-zero-2020	155	9	many	many	JJ
zkie-zero-2020	155	10	different	different	JJ
zkie-zero-2020	155	11	entities	entity	NNS
zkie-zero-2020	155	12	were	be	VBD
zkie-zero-2020	155	13	linked	link	VBN
zkie-zero-2020	155	14	to	to	IN
zkie-zero-2020	155	15	mentions	mention	NNS
zkie-zero-2020	155	16	with	with	IN
zkie-zero-2020	155	17	the	the	DT
zkie-zero-2020	155	18	same	same	JJ
zkie-zero-2020	155	19	surface	surface	NN
zkie-zero-2020	155	20	form	form	NN
zkie-zero-2020	155	21	is	be	VBZ
zkie-zero-2020	155	22	quite	quite	RB
zkie-zero-2020	155	23	high	high	JJ
zkie-zero-2020	155	24	for	for	IN
zkie-zero-2020	155	25	AIDA	AIDA	NNP
zkie-zero-2020	155	26	and	and	CC
zkie-zero-2020	155	27	WWO	WWO	NNP
zkie-zero-2020	155	28	and	and	CC
zkie-zero-2020	155	29	quite	quite	RB
zkie-zero-2020	155	30	low	low	JJ
zkie-zero-2020	155	31	for	for	IN
zkie-zero-2020	155	32	1641	1641	CD
zkie-zero-2020	155	33	.	.	.
zkie-zero-2020	156	1	We	-PRON-	PRP
zkie-zero-2020	156	2	explain	explain	VBP
zkie-zero-2020	156	3	the	the	DT
zkie-zero-2020	156	4	latter	latter	JJ
zkie-zero-2020	156	5	by	by	IN
zkie-zero-2020	156	6	the	the	DT
zkie-zero-2020	156	7	extreme	extreme	JJ
zkie-zero-2020	156	8	variance	variance	NN
zkie-zero-2020	156	9	in	in	IN
zkie-zero-2020	156	10	surface	surface	NN
zkie-zero-2020	156	11	form	form	NN
zkie-zero-2020	156	12	,	,	,
zkie-zero-2020	156	13	as	as	IN
zkie-zero-2020	156	14	even	even	RB
zkie-zero-2020	156	15	men-	men-	NN
zkie-zero-2020	156	16	tions	tion	NNS
zkie-zero-2020	156	17	of	of	IN
zkie-zero-2020	156	18	the	the	DT
zkie-zero-2020	156	19	same	same	JJ
zkie-zero-2020	156	20	name	name	NN
zkie-zero-2020	156	21	are	be	VBP
zkie-zero-2020	156	22	often	often	RB
zkie-zero-2020	156	23	written	write	VBN
zkie-zero-2020	156	24	differently	differently	RB
zkie-zero-2020	156	25	(	(	-LRB-
zkie-zero-2020	156	26	e.g.	e.g.	RB
zkie-zero-2020	157	1	Castlekevyn	Castlekevyn	NNP
zkie-zero-2020	157	2	vs.	vs.	IN
zkie-zero-2020	157	3	Castlekevin	Castlekevin	NNP
zkie-zero-2020	157	4	)	)	-RRB-
zkie-zero-2020	157	5	.	.	.
zkie-zero-2020	158	1	Also	also	RB
zkie-zero-2020	158	2	,	,	,
zkie-zero-2020	158	3	1641	1641	CD
zkie-zero-2020	158	4	contains	contain	VBZ
zkie-zero-2020	158	5	many	many	JJ
zkie-zero-2020	158	6	hapax	hapax	JJ
zkie-zero-2020	158	7	legomena	legomena	NN
zkie-zero-2020	158	8	(	(	-LRB-
zkie-zero-2020	158	9	mentions	mention	NNS
zkie-zero-2020	158	10	that	that	WDT
zkie-zero-2020	158	11	only	only	RB
zkie-zero-2020	158	12	occur	occur	VBP
zkie-zero-2020	158	13	once	once	RB
zkie-zero-2020	158	14	)	)	-RRB-
zkie-zero-2020	158	15	.	.	.
zkie-zero-2020	159	1	The	the	DT
zkie-zero-2020	159	2	average	average	JJ
zkie-zero-2020	159	3	number	number	NN
zkie-zero-2020	159	4	of	of	IN
zkie-zero-2020	159	5	candidates	candidate	NNS
zkie-zero-2020	159	6	is	be	VBZ
zkie-zero-2020	159	7	comparatively	comparatively	RB
zkie-zero-2020	159	8	larger	large	JJR
zkie-zero-2020	159	9	for	for	IN
zkie-zero-2020	159	10	WWO	WWO	NNP
zkie-zero-2020	159	11	and	and	CC
zkie-zero-2020	159	12	1641	1641	CD
zkie-zero-2020	159	13	as	as	IN
zkie-zero-2020	159	14	we	-PRON-	PRP
zkie-zero-2020	159	15	use	use	VBP
zkie-zero-2020	159	16	fuzzy	fuzzy	JJ
zkie-zero-2020	159	17	search	search	NN
zkie-zero-2020	159	18	for	for	IN
zkie-zero-2020	159	19	these	these	DT
zkie-zero-2020	159	20	.	.	.
zkie-zero-2020	160	1	Finally	finally	RB
zkie-zero-2020	160	2	,	,	,
zkie-zero-2020	160	3	the	the	DT
zkie-zero-2020	160	4	distribu-	distribu-	JJ
zkie-zero-2020	160	5	tions	tion	NNS
zkie-zero-2020	160	6	of	of	IN
zkie-zero-2020	160	7	assigned	assign	VBN
zkie-zero-2020	160	8	entities	entity	NNS
zkie-zero-2020	160	9	in	in	IN
zkie-zero-2020	160	10	WWO	WWO	NNP
zkie-zero-2020	160	11	and	and	CC
zkie-zero-2020	160	12	1641	1641	CD
zkie-zero-2020	160	13	are	be	VBP
zkie-zero-2020	160	14	also	also	RB
zkie-zero-2020	160	15	more	more	RBR
zkie-zero-2020	160	16	balanced	balanced	JJ
zkie-zero-2020	160	17	,	,	,
zkie-zero-2020	160	18	expressed	express	VBN
zkie-zero-2020	160	19	by	by	IN
zkie-zero-2020	160	20	a	a	DT
zkie-zero-2020	160	21	lower	low	JJR
zkie-zero-2020	160	22	Gini	Gini	NNP
zkie-zero-2020	160	23	co-	co-	VBD
zkie-zero-2020	160	24	efficient	efficient	JJ
zkie-zero-2020	160	25	(	(	-LRB-
zkie-zero-2020	160	26	Dodge	Dodge	NNP
zkie-zero-2020	160	27	,	,	,
zkie-zero-2020	160	28	2008	2008	CD
zkie-zero-2020	160	29	)	)	-RRB-
zkie-zero-2020	160	30	.	.	.
zkie-zero-2020	161	1	These	these	DT
zkie-zero-2020	161	2	last	last	JJ
zkie-zero-2020	161	3	two	two	CD
zkie-zero-2020	161	4	aspects	aspect	NNS
zkie-zero-2020	161	5	together	together	RB
zkie-zero-2020	161	6	with	with	IN
zkie-zero-2020	161	7	noisy	noisy	JJ
zkie-zero-2020	161	8	texts	text	NNS
zkie-zero-2020	161	9	and	and	CC
zkie-zero-2020	161	10	low	low	JJ
zkie-zero-2020	161	11	resources	resource	NNS
zkie-zero-2020	161	12	causes	cause	VBZ
zkie-zero-2020	161	13	entity	entity	NN
zkie-zero-2020	161	14	linking	link	VBG
zkie-zero-2020	161	15	to	to	TO
zkie-zero-2020	161	16	be	be	VB
zkie-zero-2020	161	17	much	much	RB
zkie-zero-2020	161	18	more	more	RBR
zkie-zero-2020	161	19	difficult	difficult	JJ
zkie-zero-2020	161	20	compared	compare	VBN
zkie-zero-2020	161	21	to	to	IN
zkie-zero-2020	161	22	state	state	NN
zkie-zero-2020	161	23	-	-	HYPH
zkie-zero-2020	161	24	of	of	IN
zkie-zero-2020	161	25	-	-	HYPH
zkie-zero-2020	161	26	the	the	DT
zkie-zero-2020	161	27	-	-	HYPH
zkie-zero-2020	161	28	art	art	NN
zkie-zero-2020	161	29	datasets	dataset	NNS
zkie-zero-2020	161	30	like	like	IN
zkie-zero-2020	161	31	AIDA	AIDA	NNP
zkie-zero-2020	161	32	.	.	.
zkie-zero-2020	162	1	5	5	CD
zkie-zero-2020	162	2	Experiments	experiment	NNS
zkie-zero-2020	162	3	To	to	TO
zkie-zero-2020	162	4	validate	validate	VB
zkie-zero-2020	162	5	our	-PRON-	PRP$
zkie-zero-2020	162	6	approach	approach	NN
zkie-zero-2020	162	7	,	,	,
zkie-zero-2020	162	8	we	-PRON-	PRP
zkie-zero-2020	162	9	first	first	RB
zkie-zero-2020	162	10	evaluate	evaluate	VBP
zkie-zero-2020	162	11	recom-	recom-	NNP
zkie-zero-2020	162	12	mender	mender	NN
zkie-zero-2020	162	13	performance	performance	NN
zkie-zero-2020	162	14	.	.	.
zkie-zero-2020	163	1	Then	then	RB
zkie-zero-2020	163	2	,	,	,
zkie-zero-2020	163	3	non	non	JJ
zkie-zero-2020	163	4	-	-	JJ
zkie-zero-2020	163	5	interactive	interactive	JJ
zkie-zero-2020	163	6	rank-	rank-	NN
zkie-zero-2020	163	7	ing	ing	NN
zkie-zero-2020	163	8	performance	performance	NN
zkie-zero-2020	163	9	is	be	VBZ
zkie-zero-2020	163	10	evaluated	evaluate	VBN
zkie-zero-2020	163	11	similarly	similarly	RB
zkie-zero-2020	163	12	to	to	IN
zkie-zero-2020	163	13	state	state	NN
zkie-zero-2020	163	14	-	-	HYPH
zkie-zero-2020	163	15	of-	of-	VBN
zkie-zero-2020	163	16	the	the	DT
zkie-zero-2020	163	17	-	-	HYPH
zkie-zero-2020	163	18	art	art	NN
zkie-zero-2020	163	19	EL	el	NN
zkie-zero-2020	163	20	.	.	.
zkie-zero-2020	164	1	Afterwards	afterwards	RB
zkie-zero-2020	164	2	,	,	,
zkie-zero-2020	164	3	we	-PRON-	PRP
zkie-zero-2020	164	4	simulate	simulate	VBP
zkie-zero-2020	164	5	a	a	DT
zkie-zero-2020	164	6	user	user	NN
zkie-zero-2020	164	7	annotat-	annotat-	NN
zkie-zero-2020	164	8	ing	e	VBG
zkie-zero-2020	164	9	corpora	corpora	NN
zkie-zero-2020	164	10	with	with	IN
zkie-zero-2020	164	11	our	-PRON-	PRP$
zkie-zero-2020	164	12	Human	human	JJ
zkie-zero-2020	164	13	-	-	HYPH
zkie-zero-2020	164	14	In	in	IN
zkie-zero-2020	164	15	-	-	HYPH
zkie-zero-2020	164	16	The	The	NNP
zkie-zero-2020	164	17	-	-	HYPH
zkie-zero-2020	164	18	Loop	Loop	NNP
zkie-zero-2020	164	19	ranker	ranker	NN
zkie-zero-2020	164	20	.	.	.
zkie-zero-2020	165	1	Finally	finally	RB
zkie-zero-2020	165	2	,	,	,
zkie-zero-2020	165	3	we	-PRON-	PRP
zkie-zero-2020	165	4	conduct	conduct	VBP
zkie-zero-2020	165	5	a	a	DT
zkie-zero-2020	165	6	user	user	NN
zkie-zero-2020	165	7	study	study	NN
zkie-zero-2020	165	8	to	to	TO
zkie-zero-2020	165	9	test	test	VB
zkie-zero-2020	165	10	it	-PRON-	PRP
zkie-zero-2020	165	11	in	in	IN
zkie-zero-2020	165	12	a	a	DT
zkie-zero-2020	165	13	re-	re-	RB
zkie-zero-2020	165	14	alistic	alistic	JJ
zkie-zero-2020	165	15	setting	setting	NN
zkie-zero-2020	165	16	.	.	.
zkie-zero-2020	166	1	Similar	similar	JJ
zkie-zero-2020	166	2	to	to	IN
zkie-zero-2020	166	3	other	other	JJ
zkie-zero-2020	166	4	work	work	NN
zkie-zero-2020	166	5	on	on	IN
zkie-zero-2020	166	6	EL	el	NN
zkie-zero-2020	166	7	,	,	,
zkie-zero-2020	166	8	our	-PRON-	PRP$
zkie-zero-2020	166	9	main	main	JJ
zkie-zero-2020	166	10	metric	metric	NN
zkie-zero-2020	166	11	for	for	IN
zkie-zero-2020	166	12	ranking	ranking	NN
zkie-zero-2020	166	13	is	be	VBZ
zkie-zero-2020	166	14	accuracy	accuracy	NN
zkie-zero-2020	166	15	.	.	.
zkie-zero-2020	167	1	We	-PRON-	PRP
zkie-zero-2020	167	2	also	also	RB
zkie-zero-2020	167	3	mea-	mea-	VBZ
zkie-zero-2020	167	4	sure	sure	JJ
zkie-zero-2020	167	5	Accuracy@5	Accuracy@5	JJS
zkie-zero-2020	167	6	,	,	,
zkie-zero-2020	167	7	as	as	IN
zkie-zero-2020	167	8	our	-PRON-	PRP$
zkie-zero-2020	167	9	experiments	experiment	NNS
zkie-zero-2020	167	10	showed	show	VBD
zkie-zero-2020	167	11	that	that	IN
zkie-zero-2020	167	12	users	user	NNS
zkie-zero-2020	167	13	can	can	MD
zkie-zero-2020	167	14	quickly	quickly	RB
zkie-zero-2020	167	15	scan	scan	VB
zkie-zero-2020	167	16	and	and	CC
zkie-zero-2020	167	17	select	select	VB
zkie-zero-2020	167	18	the	the	DT
zkie-zero-2020	167	19	right	right	JJ
zkie-zero-2020	167	20	entity	entity	NN
zkie-zero-2020	167	21	from	from	IN
zkie-zero-2020	167	22	a	a	DT
zkie-zero-2020	167	23	list	list	NN
zkie-zero-2020	167	24	of	of	IN
zkie-zero-2020	167	25	five	five	CD
zkie-zero-2020	167	26	elements	element	NNS
zkie-zero-2020	167	27	.	.	.
zkie-zero-2020	168	1	In	in	IN
zkie-zero-2020	168	2	our	-PRON-	PRP$
zkie-zero-2020	168	3	annotation	annotation	NN
zkie-zero-2020	168	4	edi-	edi-	NN
zkie-zero-2020	168	5	tor	tor	NN
zkie-zero-2020	168	6	,	,	,
zkie-zero-2020	168	7	the	the	DT
zkie-zero-2020	168	8	candidate	candidate	NN
zkie-zero-2020	168	9	list	list	NN
zkie-zero-2020	168	10	shows	show	VBZ
zkie-zero-2020	168	11	the	the	DT
zkie-zero-2020	168	12	first	first	JJ
zkie-zero-2020	168	13	five	five	CD
zkie-zero-2020	168	14	elements	element	NNS
zkie-zero-2020	168	15	without	without	IN
zkie-zero-2020	168	16	scrolling	scroll	VBG
zkie-zero-2020	168	17	.	.	.
zkie-zero-2020	169	1	As	as	IN
zkie-zero-2020	169	2	a	a	DT
zkie-zero-2020	169	3	baseline	baseline	NN
zkie-zero-2020	169	4	,	,	,
zkie-zero-2020	169	5	we	-PRON-	PRP
zkie-zero-2020	169	6	use	use	VBP
zkie-zero-2020	169	7	the	the	DT
zkie-zero-2020	169	8	Most-	most-	NN
zkie-zero-2020	169	9	Frequently	frequently	RB
zkie-zero-2020	169	10	Linked	link	VBD
zkie-zero-2020	169	11	Entity	entity	NN
zkie-zero-2020	169	12	baseline	baseline	NN
zkie-zero-2020	169	13	(	(	-LRB-
zkie-zero-2020	169	14	MFLEB	MFLEB	NNP
zkie-zero-2020	169	15	)	)	-RRB-
zkie-zero-2020	169	16	.	.	.
zkie-zero-2020	170	1	It	-PRON-	PRP
zkie-zero-2020	170	2	assigns	assign	VBZ
zkie-zero-2020	170	3	,	,	,
zkie-zero-2020	170	4	given	give	VBN
zkie-zero-2020	170	5	a	a	DT
zkie-zero-2020	170	6	mention	mention	NN
zkie-zero-2020	170	7	,	,	,
zkie-zero-2020	170	8	the	the	DT
zkie-zero-2020	170	9	entity	entity	NN
zkie-zero-2020	170	10	that	that	WDT
zkie-zero-2020	170	11	was	be	VBD
zkie-zero-2020	170	12	most	most	RBS
zkie-zero-2020	170	13	often	often	RB
zkie-zero-2020	170	14	linked	link	VBN
zkie-zero-2020	170	15	to	to	IN
zkie-zero-2020	170	16	it	-PRON-	PRP
zkie-zero-2020	170	17	in	in	IN
zkie-zero-2020	170	18	the	the	DT
zkie-zero-2020	170	19	training	training	NN
zkie-zero-2020	170	20	data	datum	NNS
zkie-zero-2020	170	21	.	.	.
zkie-zero-2020	171	1	5.1	5.1	CD
zkie-zero-2020	171	2	Automatic	automatic	JJ
zkie-zero-2020	171	3	suggestion	suggestion	NN
zkie-zero-2020	171	4	performance	performance	NN
zkie-zero-2020	171	5	We	-PRON-	PRP
zkie-zero-2020	171	6	evaluate	evaluate	VBP
zkie-zero-2020	171	7	the	the	DT
zkie-zero-2020	171	8	performance	performance	NN
zkie-zero-2020	171	9	of	of	IN
zkie-zero-2020	171	10	our	-PRON-	PRP$
zkie-zero-2020	171	11	Levenshtein-	Levenshtein-	NNP
zkie-zero-2020	171	12	based	base	VBN
zkie-zero-2020	171	13	recommender	recommender	NN
zkie-zero-2020	171	14	that	that	WDT
zkie-zero-2020	171	15	suggests	suggest	VBZ
zkie-zero-2020	171	16	potential	potential	JJ
zkie-zero-2020	171	17	annota-	annota-	JJ
zkie-zero-2020	171	18	tions	tion	NNS
zkie-zero-2020	171	19	to	to	IN
zkie-zero-2020	171	20	users	user	NNS
zkie-zero-2020	171	21	(	(	-LRB-
zkie-zero-2020	171	22	Table	table	NN
zkie-zero-2020	171	23	3	3	CD
zkie-zero-2020	171	24	)	)	-RRB-
zkie-zero-2020	171	25	.	.	.
zkie-zero-2020	172	1	We	-PRON-	PRP
zkie-zero-2020	172	2	filter	filter	VBP
zkie-zero-2020	172	3	out	out	RP
zkie-zero-2020	172	4	suggestions	suggestion	NNS
zkie-zero-2020	172	5	consisting	consist	VBG
zkie-zero-2020	172	6	of	of	IN
zkie-zero-2020	172	7	≤	≤	JJ
zkie-zero-2020	172	8	3	3	CD
zkie-zero-2020	172	9	characters	character	NNS
zkie-zero-2020	172	10	as	as	IN
zkie-zero-2020	172	11	these	these	DT
zkie-zero-2020	172	12	introduce	introduce	VBP
zkie-zero-2020	172	13	too	too	RB
zkie-zero-2020	172	14	much	much	JJ
zkie-zero-2020	172	15	noise	noise	NN
zkie-zero-2020	172	16	.	.	.
zkie-zero-2020	173	1	For	for	IN
zkie-zero-2020	173	2	annotation	annotation	NN
zkie-zero-2020	173	3	suggestions	suggestion	NNS
zkie-zero-2020	173	4	,	,	,
zkie-zero-2020	173	5	we	-PRON-	PRP
zkie-zero-2020	173	6	focus	focus	VBP
zkie-zero-2020	173	7	on	on	IN
zkie-zero-2020	173	8	recall	recall	NN
zkie-zero-2020	173	9	:	:	:
zkie-zero-2020	173	10	where	where	WRB
zkie-zero-2020	173	11	low	low	JJ
zkie-zero-2020	173	12	precision	precision	NN
zkie-zero-2020	173	13	implies	imply	VBZ
zkie-zero-2020	173	14	rec-	rec-	NN
zkie-zero-2020	173	15	ommendations	ommendation	NNS
zkie-zero-2020	173	16	that	that	WDT
zkie-zero-2020	173	17	are	be	VBP
zkie-zero-2020	173	18	not	not	RB
zkie-zero-2020	173	19	useful	useful	JJ
zkie-zero-2020	173	20	,	,	,
zkie-zero-2020	173	21	no	no	DT
zkie-zero-2020	173	22	recall	recall	NN
zkie-zero-2020	173	23	results	result	NNS
zkie-zero-2020	173	24	in	in	IN
zkie-zero-2020	173	25	no	no	DT
zkie-zero-2020	173	26	recommendations	recommendation	NNS
zkie-zero-2020	173	27	at	at	RB
zkie-zero-2020	173	28	all	all	RB
zkie-zero-2020	173	29	.	.	.
zkie-zero-2020	174	1	It	-PRON-	PRP
zkie-zero-2020	174	2	can	can	MD
zkie-zero-2020	174	3	be	be	VB
zkie-zero-2020	174	4	seen	see	VBN
zkie-zero-2020	174	5	that	that	IN
zkie-zero-2020	174	6	for	for	IN
zkie-zero-2020	174	7	AIDA	AIDA	NNP
zkie-zero-2020	174	8	and	and	CC
zkie-zero-2020	174	9	WWO	WWO	NNP
zkie-zero-2020	174	10	,	,	,
zkie-zero-2020	174	11	the	the	DT
zkie-zero-2020	174	12	performance	performance	NN
zkie-zero-2020	174	13	of	of	IN
zkie-zero-2020	174	14	all	all	DT
zkie-zero-2020	174	15	three	three	CD
zkie-zero-2020	174	16	recommenders	recommender	NNS
zkie-zero-2020	174	17	is	be	VBZ
zkie-zero-2020	174	18	quite	quite	RB
zkie-zero-2020	174	19	good	good	JJ
zkie-zero-2020	174	20	(	(	-LRB-
zkie-zero-2020	174	21	recall	recall	NN
zkie-zero-2020	174	22	is	be	VBZ
zkie-zero-2020	174	23	about	about	RB
zkie-zero-2020	174	24	60	60	CD
zkie-zero-2020	174	25	%	%	NN
zkie-zero-2020	174	26	and	and	CC
zkie-zero-2020	174	27	40	40	CD
zkie-zero-2020	174	28	%	%	NN
zkie-zero-2020	174	29	)	)	-RRB-
zkie-zero-2020	174	30	while	while	IN
zkie-zero-2020	174	31	for	for	IN
zkie-zero-2020	174	32	1641	1641	CD
zkie-zero-2020	174	33	,	,	,
zkie-zero-2020	174	34	it	-PRON-	PRP
zkie-zero-2020	174	35	is	be	VBZ
zkie-zero-2020	174	36	only	only	RB
zkie-zero-2020	174	37	around	around	IN
zkie-zero-2020	174	38	20	20	CD
zkie-zero-2020	174	39	%	%	NN
zkie-zero-2020	174	40	.	.	.
zkie-zero-2020	175	1	The	the	DT
zkie-zero-2020	175	2	Levenshtein	Levenshtein	NNP
zkie-zero-2020	175	3	recommender	recommender	NN
zkie-zero-2020	175	4	increases	increase	VBZ
zkie-zero-2020	175	5	recall	recall	VBP
zkie-zero-2020	175	6	and	and	CC
zkie-zero-2020	175	7	reduces	reduce	VBZ
zkie-zero-2020	175	8	precision	precision	NN
zkie-zero-2020	175	9	.	.	.
zkie-zero-2020	176	1	The	the	DT
zkie-zero-2020	176	2	impact	impact	NN
zkie-zero-2020	176	3	is	be	VBZ
zkie-zero-2020	176	4	most	most	RBS
zkie-zero-2020	176	5	pronounced	pronounce	VBN
zkie-zero-2020	176	6	for	for	IN
zkie-zero-2020	176	7	1641	1641	CD
zkie-zero-2020	176	8	,	,	,
zkie-zero-2020	176	9	where	where	WRB
zkie-zero-2020	176	10	it	-PRON-	PRP
zkie-zero-2020	176	11	improves	improve	VBZ
zkie-zero-2020	176	12	recall	recall	NN
zkie-zero-2020	176	13	upon	upon	IN
zkie-zero-2020	176	14	the	the	DT
zkie-zero-2020	176	15	string	string	NN
zkie-zero-2020	176	16	matching	match	VBG
zkie-zero-2020	176	17	recommender	recommender	NN
zkie-zero-2020	176	18	by	by	IN
zkie-zero-2020	176	19	around	around	RB
zkie-zero-2020	176	20	50	50	CD
zkie-zero-2020	176	21	%	%	NN
zkie-zero-2020	176	22	.	.	.
zkie-zero-2020	177	1	In	in	IN
zkie-zero-2020	177	2	sum-	sum-	NNP
zkie-zero-2020	177	3	mary	mary	NNP
zkie-zero-2020	177	4	,	,	,
zkie-zero-2020	177	5	we	-PRON-	PRP
zkie-zero-2020	177	6	suggest	suggest	VBP
zkie-zero-2020	177	7	using	use	VBG
zkie-zero-2020	177	8	the	the	DT
zkie-zero-2020	177	9	string	string	NN
zkie-zero-2020	177	10	matching	match	VBG
zkie-zero-2020	177	11	rec-	rec-	NN
zkie-zero-2020	177	12	6987	6987	CD
zkie-zero-2020	177	13	Dataset	Dataset	NNP
zkie-zero-2020	177	14	Model	Model	NNP
zkie-zero-2020	177	15	P	P	NNP
zkie-zero-2020	177	16	R	R	NNP
zkie-zero-2020	177	17	F1	f1	NN
zkie-zero-2020	177	18	AIDA	AIDA	NNP
zkie-zero-2020	177	19	String	String	NNP
zkie-zero-2020	177	20	0.43	0.43	CD
zkie-zero-2020	177	21	0.60	0.60	CD
zkie-zero-2020	177	22	0.50	0.50	CD
zkie-zero-2020	177	23	Leven@1	leven@1	CD
zkie-zero-2020	177	24	0.31	0.31	CD
zkie-zero-2020	177	25	0.55	0.55	CD
zkie-zero-2020	177	26	0.40	0.40	CD
zkie-zero-2020	177	27	Leven@2	leven@2	CD
zkie-zero-2020	177	28	0.19	0.19	CD
zkie-zero-2020	177	29	0.57	0.57	CD
zkie-zero-2020	177	30	0.28	0.28	CD
zkie-zero-2020	177	31	WWO	WWO	NNP
zkie-zero-2020	177	32	String	String	NNP
zkie-zero-2020	177	33	0.17	0.17	CD
zkie-zero-2020	177	34	0.38	0.38	CD
zkie-zero-2020	177	35	0.23	0.23	CD
zkie-zero-2020	177	36	Leven@1	leven@1	CD
zkie-zero-2020	177	37	0.11	0.11	CD
zkie-zero-2020	177	38	0.40	0.40	CD
zkie-zero-2020	177	39	0.16	0.16	CD
zkie-zero-2020	177	40	Leven@2	leven@2	CD
zkie-zero-2020	177	41	0.04	0.04	CD
zkie-zero-2020	177	42	0.42	0.42	CD
zkie-zero-2020	177	43	0.07	0.07	CD
zkie-zero-2020	177	44	1641	1641	CD
zkie-zero-2020	177	45	String	String	NNP
zkie-zero-2020	177	46	0.12	0.12	CD
zkie-zero-2020	177	47	0.14	0.14	CD
zkie-zero-2020	177	48	0.13	0.13	CD
zkie-zero-2020	177	49	Leven@1	leven@1	CD
zkie-zero-2020	177	50	0.16	0.16	CD
zkie-zero-2020	177	51	0.19	0.19	CD
zkie-zero-2020	177	52	0.17	0.17	CD
zkie-zero-2020	177	53	Leven@2	leven@2	$
zkie-zero-2020	177	54	0.12	0.12	CD
zkie-zero-2020	177	55	0.22	0.22	CD
zkie-zero-2020	177	56	0.15	0.15	CD
zkie-zero-2020	177	57	Table	table	NN
zkie-zero-2020	177	58	3	3	CD
zkie-zero-2020	177	59	:	:	:
zkie-zero-2020	177	60	Recommender	recommender	NN
zkie-zero-2020	177	61	performance	performance	NN
zkie-zero-2020	177	62	in	in	IN
zkie-zero-2020	177	63	Precision	Precision	NNP
zkie-zero-2020	177	64	,	,	,
zkie-zero-2020	177	65	Recall	Recall	NNP
zkie-zero-2020	177	66	and	and	CC
zkie-zero-2020	177	67	F1	f1	NN
zkie-zero-2020	177	68	score	score	NN
zkie-zero-2020	177	69	for	for	IN
zkie-zero-2020	177	70	String	string	NN
zkie-zero-2020	177	71	matching	match	VBG
zkie-zero-2020	177	72	recommender	recommender	NN
zkie-zero-2020	177	73	and	and	CC
zkie-zero-2020	177	74	Levenshtein	Levenshtein	NNP
zkie-zero-2020	177	75	recommender	recommender	NN
zkie-zero-2020	177	76	with	with	IN
zkie-zero-2020	177	77	distance	distance	NN
zkie-zero-2020	177	78	1	1	CD
zkie-zero-2020	177	79	and	and	CC
zkie-zero-2020	177	80	2	2	CD
zkie-zero-2020	177	81	.	.	.
zkie-zero-2020	178	1	For	for	IN
zkie-zero-2020	178	2	AIDA	AIDA	NNP
zkie-zero-2020	178	3	,	,	,
zkie-zero-2020	178	4	we	-PRON-	PRP
zkie-zero-2020	178	5	evaluate	evaluate	VBP
zkie-zero-2020	178	6	on	on	IN
zkie-zero-2020	178	7	the	the	DT
zkie-zero-2020	178	8	test	test	NN
zkie-zero-2020	178	9	set	set	NN
zkie-zero-2020	178	10	,	,	,
zkie-zero-2020	178	11	for	for	IN
zkie-zero-2020	178	12	the	the	DT
zkie-zero-2020	178	13	other	other	JJ
zkie-zero-2020	178	14	datasets	dataset	NNS
zkie-zero-2020	178	15	,	,	,
zkie-zero-2020	178	16	we	-PRON-	PRP
zkie-zero-2020	178	17	use	use	VBP
zkie-zero-2020	178	18	10-fold	10-fold	CD
zkie-zero-2020	178	19	cross	cross	NN
zkie-zero-2020	178	20	validation	validation	NN
zkie-zero-2020	178	21	.	.	.
zkie-zero-2020	179	1	ommender	ommender	VB
zkie-zero-2020	179	2	for	for	IN
zkie-zero-2020	179	3	domains	domain	NNS
zkie-zero-2020	179	4	where	where	WRB
zkie-zero-2020	179	5	texts	text	NNS
zkie-zero-2020	179	6	are	be	VBP
zkie-zero-2020	179	7	clean	clean	JJ
zkie-zero-2020	179	8	and	and	CC
zkie-zero-2020	179	9	exhibit	exhibit	VB
zkie-zero-2020	179	10	low	low	JJ
zkie-zero-2020	179	11	spelling	spelling	NN
zkie-zero-2020	179	12	variance	variance	NN
zkie-zero-2020	179	13	.	.	.
zkie-zero-2020	180	1	We	-PRON-	PRP
zkie-zero-2020	180	2	consider	consider	VBP
zkie-zero-2020	180	3	the	the	DT
zkie-zero-2020	180	4	Levenshtein	Levenshtein	NNP
zkie-zero-2020	180	5	recommender	recommender	NN
zkie-zero-2020	180	6	to	to	TO
zkie-zero-2020	180	7	be	be	VB
zkie-zero-2020	180	8	more	more	RBR
zkie-zero-2020	180	9	suitable	suitable	JJ
zkie-zero-2020	180	10	for	for	IN
zkie-zero-2020	180	11	domains	domain	NNS
zkie-zero-2020	180	12	with	with	IN
zkie-zero-2020	180	13	noisy	noisy	JJ
zkie-zero-2020	180	14	texts	text	NNS
zkie-zero-2020	180	15	.	.	.
zkie-zero-2020	181	1	5.2	5.2	CD
zkie-zero-2020	181	2	Candidate	candidate	NN
zkie-zero-2020	181	3	ranking	rank	VBG
zkie-zero-2020	181	4	performance	performance	NN
zkie-zero-2020	181	5	We	-PRON-	PRP
zkie-zero-2020	181	6	evaluate	evaluate	VBP
zkie-zero-2020	181	7	EL	el	NN
zkie-zero-2020	181	8	candidate	candidate	NN
zkie-zero-2020	181	9	ranking	rank	VBG
zkie-zero-2020	181	10	in	in	IN
zkie-zero-2020	181	11	a	a	DT
zkie-zero-2020	181	12	non-	non-	NNP
zkie-zero-2020	181	13	interactive	interactive	JJ
zkie-zero-2020	181	14	setting	set	VBG
zkie-zero-2020	181	15	first	first	RB
zkie-zero-2020	181	16	to	to	TO
zkie-zero-2020	181	17	estimate	estimate	VB
zkie-zero-2020	181	18	the	the	DT
zkie-zero-2020	181	19	upper	upper	JJ
zkie-zero-2020	181	20	bound	bind	VBN
zkie-zero-2020	181	21	ranking	rank	VBG
zkie-zero-2020	181	22	performance	performance	NN
zkie-zero-2020	181	23	.	.	.
zkie-zero-2020	182	1	As	as	IN
zkie-zero-2020	182	2	we	-PRON-	PRP
zkie-zero-2020	182	3	are	be	VBP
zkie-zero-2020	182	4	the	the	DT
zkie-zero-2020	182	5	first	first	JJ
zkie-zero-2020	182	6	to	to	IN
zkie-zero-2020	182	7	per-	per-	DT
zkie-zero-2020	182	8	form	form	VB
zkie-zero-2020	182	9	EL	el	NN
zkie-zero-2020	182	10	on	on	IN
zkie-zero-2020	182	11	our	-PRON-	PRP$
zkie-zero-2020	182	12	version	version	NN
zkie-zero-2020	182	13	of	of	IN
zkie-zero-2020	182	14	WWO	WWO	NNP
zkie-zero-2020	182	15	and	and	CC
zkie-zero-2020	182	16	1641	1641	CD
zkie-zero-2020	182	17	,	,	,
zkie-zero-2020	182	18	it	-PRON-	PRP
zkie-zero-2020	182	19	also	also	RB
zkie-zero-2020	182	20	serves	serve	VBZ
zkie-zero-2020	182	21	as	as	IN
zkie-zero-2020	182	22	a	a	DT
zkie-zero-2020	182	23	difficulty	difficulty	NN
zkie-zero-2020	182	24	comparison	comparison	NN
zkie-zero-2020	182	25	between	between	IN
zkie-zero-2020	182	26	AIDA	AIDA	NNP
zkie-zero-2020	182	27	as	as	IN
zkie-zero-2020	182	28	the	the	DT
zkie-zero-2020	182	29	state	state	NN
zkie-zero-2020	182	30	-	-	HYPH
zkie-zero-2020	182	31	of	of	IN
zkie-zero-2020	182	32	-	-	HYPH
zkie-zero-2020	182	33	the	the	DT
zkie-zero-2020	182	34	-	-	HYPH
zkie-zero-2020	182	35	art	art	NN
zkie-zero-2020	182	36	dataset	dataset	NN
zkie-zero-2020	182	37	and	and	CC
zkie-zero-2020	182	38	datasets	dataset	VBZ
zkie-zero-2020	182	39	from	from	IN
zkie-zero-2020	182	40	our	-PRON-	PRP$
zkie-zero-2020	182	41	domain	domain	NN
zkie-zero-2020	182	42	-	-	HYPH
zkie-zero-2020	182	43	specific	specific	JJ
zkie-zero-2020	182	44	setting	setting	NN
zkie-zero-2020	182	45	.	.	.
zkie-zero-2020	183	1	For	for	IN
zkie-zero-2020	183	2	AIDA	AIDA	NNP
zkie-zero-2020	183	3	,	,	,
zkie-zero-2020	183	4	we	-PRON-	PRP
zkie-zero-2020	183	5	use	use	VBP
zkie-zero-2020	183	6	the	the	DT
zkie-zero-2020	183	7	ex-	ex-	NN
zkie-zero-2020	183	8	isting	iste	VBG
zkie-zero-2020	183	9	train	train	NN
zkie-zero-2020	183	10	,	,	,
zkie-zero-2020	183	11	development	development	NN
zkie-zero-2020	183	12	and	and	CC
zkie-zero-2020	183	13	test	test	NN
zkie-zero-2020	183	14	split	split	NN
zkie-zero-2020	183	15	;	;	:
zkie-zero-2020	183	16	for	for	IN
zkie-zero-2020	183	17	the	the	DT
zkie-zero-2020	183	18	other	other	JJ
zkie-zero-2020	183	19	two	two	CD
zkie-zero-2020	183	20	corpora	corpora	NN
zkie-zero-2020	183	21	,	,	,
zkie-zero-2020	183	22	we	-PRON-	PRP
zkie-zero-2020	183	23	perform	perform	VBP
zkie-zero-2020	183	24	10-fold	10-fold	CD
zkie-zero-2020	183	25	cross	cross	NN
zkie-zero-2020	183	26	validation	validation	NN
zkie-zero-2020	183	27	as	as	IN
zkie-zero-2020	183	28	we	-PRON-	PRP
zkie-zero-2020	183	29	observed	observe	VBD
zkie-zero-2020	183	30	high	high	JJ
zkie-zero-2020	183	31	variance	variance	NN
zkie-zero-2020	183	32	in	in	IN
zkie-zero-2020	183	33	score	score	NN
zkie-zero-2020	183	34	when	when	WRB
zkie-zero-2020	183	35	us-	us-	XX
zkie-zero-2020	183	36	ing	e	VBG
zkie-zero-2020	183	37	different	different	JJ
zkie-zero-2020	183	38	train	train	NN
zkie-zero-2020	183	39	-	-	HYPH
zkie-zero-2020	183	40	test	test	NN
zkie-zero-2020	183	41	splits	split	NNS
zkie-zero-2020	183	42	.	.	.
zkie-zero-2020	184	1	Features	feature	NNS
zkie-zero-2020	184	2	related	relate	VBN
zkie-zero-2020	184	3	to	to	IN
zkie-zero-2020	184	4	user	user	NN
zkie-zero-2020	184	5	queries	query	NNS
zkie-zero-2020	184	6	are	be	VBP
zkie-zero-2020	184	7	not	not	RB
zkie-zero-2020	184	8	used	use	VBN
zkie-zero-2020	184	9	in	in	IN
zkie-zero-2020	184	10	this	this	DT
zkie-zero-2020	184	11	experiment	experiment	NN
zkie-zero-2020	184	12	.	.	.
zkie-zero-2020	185	1	We	-PRON-	PRP
zkie-zero-2020	185	2	assume	assume	VBP
zkie-zero-2020	185	3	that	that	IN
zkie-zero-2020	185	4	the	the	DT
zkie-zero-2020	185	5	gold	gold	JJ
zkie-zero-2020	185	6	candidate	candidate	NN
zkie-zero-2020	185	7	always	always	RB
zkie-zero-2020	185	8	exists	exist	VBZ
zkie-zero-2020	185	9	in	in	IN
zkie-zero-2020	185	10	training	training	NN
zkie-zero-2020	185	11	and	and	CC
zkie-zero-2020	185	12	evaluation	evaluation	NN
zkie-zero-2020	185	13	data	datum	NNS
zkie-zero-2020	185	14	.	.	.
zkie-zero-2020	186	1	The	the	DT
zkie-zero-2020	186	2	results	result	NNS
zkie-zero-2020	186	3	of	of	IN
zkie-zero-2020	186	4	this	this	DT
zkie-zero-2020	186	5	experiment	experiment	NN
zkie-zero-2020	186	6	are	be	VBP
zkie-zero-2020	186	7	depicted	depict	VBN
zkie-zero-2020	186	8	in	in	IN
zkie-zero-2020	186	9	Table	table	NN
zkie-zero-2020	186	10	4	4	CD
zkie-zero-2020	186	11	.	.	.
zkie-zero-2020	187	1	It	-PRON-	PRP
zkie-zero-2020	187	2	can	can	MD
zkie-zero-2020	187	3	be	be	VB
zkie-zero-2020	187	4	seen	see	VBN
zkie-zero-2020	187	5	that	that	IN
zkie-zero-2020	187	6	for	for	IN
zkie-zero-2020	187	7	AIDA	AIDA	NNP
zkie-zero-2020	187	8	,	,	,
zkie-zero-2020	187	9	the	the	DT
zkie-zero-2020	187	10	MFLE	MFLE	NNP
zkie-zero-2020	187	11	baseline	baseline	NN
zkie-zero-2020	187	12	is	be	VBZ
zkie-zero-2020	187	13	particularly	particularly	RB
zkie-zero-2020	187	14	strong	strong	JJ
zkie-zero-2020	187	15	,	,	,
zkie-zero-2020	187	16	being	be	VBG
zkie-zero-2020	187	17	better	well	JJR
zkie-zero-2020	187	18	than	than	IN
zkie-zero-2020	187	19	all	all	DT
zkie-zero-2020	187	20	trained	train	VBN
zkie-zero-2020	187	21	models	model	NNS
zkie-zero-2020	187	22	.	.	.
zkie-zero-2020	188	1	For	for	IN
zkie-zero-2020	188	2	the	the	DT
zkie-zero-2020	188	3	other	other	JJ
zkie-zero-2020	188	4	datasets	dataset	NNS
zkie-zero-2020	188	5	,	,	,
zkie-zero-2020	188	6	the	the	DT
zkie-zero-2020	188	7	baseline	baseline	NN
zkie-zero-2020	188	8	is	be	VBZ
zkie-zero-2020	188	9	weaker	weak	JJR
zkie-zero-2020	188	10	than	than	IN
zkie-zero-2020	188	11	all	all	DT
zkie-zero-2020	188	12	,	,	,
zkie-zero-2020	188	13	show-	show-	VBG
zkie-zero-2020	188	14	ing	e	VBG
zkie-zero-2020	188	15	that	that	DT
zkie-zero-2020	188	16	popularity	popularity	NN
zkie-zero-2020	188	17	is	be	VBZ
zkie-zero-2020	188	18	a	a	DT
zkie-zero-2020	188	19	weak	weak	JJ
zkie-zero-2020	188	20	feature	feature	NN
zkie-zero-2020	188	21	in	in	IN
zkie-zero-2020	188	22	our	-PRON-	PRP$
zkie-zero-2020	188	23	setting	setting	NN
zkie-zero-2020	188	24	.	.	.
zkie-zero-2020	189	1	For	for	IN
zkie-zero-2020	189	2	AIDA	AIDA	NNP
zkie-zero-2020	189	3	,	,	,
zkie-zero-2020	189	4	LightGBM	LightGBM	NNP
zkie-zero-2020	189	5	performs	perform	VBZ
zkie-zero-2020	189	6	best	well	RBS
zkie-zero-2020	189	7	,	,	,
zkie-zero-2020	189	8	for	for	IN
zkie-zero-2020	189	9	WWO	WWO	NNP
zkie-zero-2020	189	10	and	and	CC
zkie-zero-2020	189	11	1641	1641	CD
zkie-zero-2020	189	12	,	,	,
zkie-zero-2020	189	13	the	the	DT
zkie-zero-2020	189	14	RankNet	RankNet	NNP
zkie-zero-2020	189	15	is	be	VBZ
zkie-zero-2020	189	16	best	well	RBS
zkie-zero-2020	189	17	closely	closely	RB
zkie-zero-2020	189	18	followed	follow	VBN
zkie-zero-2020	189	19	by	by	IN
zkie-zero-2020	189	20	the	the	DT
zkie-zero-2020	189	21	RankSVM	ranksvm	NN
zkie-zero-2020	189	22	.	.	.
zkie-zero-2020	190	1	The	the	DT
zkie-zero-2020	190	2	accuracy@5	accuracy@5	CD
zkie-zero-2020	190	3	is	be	VBZ
zkie-zero-2020	190	4	compara-	compara-	XX
zkie-zero-2020	190	5	tively	tively	RB
zkie-zero-2020	190	6	high	high	RB
zkie-zero-2020	190	7	as	as	IN
zkie-zero-2020	190	8	there	there	EX
zkie-zero-2020	190	9	are	be	VBP
zkie-zero-2020	190	10	cases	case	NNS
zkie-zero-2020	190	11	where	where	WRB
zkie-zero-2020	190	12	the	the	DT
zkie-zero-2020	190	13	candidate	candidate	NN
zkie-zero-2020	190	14	list	list	NN
zkie-zero-2020	190	15	is	be	VBZ
zkie-zero-2020	190	16	relatively	relatively	RB
zkie-zero-2020	190	17	short	short	JJ
zkie-zero-2020	190	18	.	.	.
zkie-zero-2020	191	1	Regarding	regard	VBG
zkie-zero-2020	191	2	training	training	NN
zkie-zero-2020	191	3	times	time	NNS
zkie-zero-2020	191	4	,	,	,
zkie-zero-2020	191	5	LightGBM	LightGBM	NNP
zkie-zero-2020	191	6	trains	train	NNS
zkie-zero-2020	191	7	extremely	extremely	RB
zkie-zero-2020	191	8	fast	fast	VBP
zkie-zero-2020	191	9	with	with	IN
zkie-zero-2020	191	10	RankSVM	RankSVM	.
zkie-zero-2020	191	11	being	be	VBG
zkie-zero-2020	191	12	a	a	DT
zkie-zero-2020	191	13	close	close	JJ
zkie-zero-2020	191	14	second	second	NN
zkie-zero-2020	191	15	.	.	.
zkie-zero-2020	192	1	They	-PRON-	PRP
zkie-zero-2020	192	2	are	be	VBP
zkie-zero-2020	192	3	fast	fast	JJ
zkie-zero-2020	192	4	enough	enough	RB
zkie-zero-2020	192	5	to	to	TO
zkie-zero-2020	192	6	re-	re-	RB
zkie-zero-2020	192	7	train	train	VB
zkie-zero-2020	192	8	after	after	IN
zkie-zero-2020	192	9	each	each	DT
zkie-zero-2020	192	10	user	user	NN
zkie-zero-2020	192	11	annotation	annotation	NN
zkie-zero-2020	192	12	.	.	.
zkie-zero-2020	193	1	The	the	DT
zkie-zero-2020	193	2	RankNet	RankNet	NNP
zkie-zero-2020	193	3	trains	train	NNS
zkie-zero-2020	193	4	two	two	CD
zkie-zero-2020	193	5	to	to	TO
zkie-zero-2020	193	6	four	four	CD
zkie-zero-2020	193	7	times	time	NNS
zkie-zero-2020	193	8	slower	slow	JJR
zkie-zero-2020	193	9	than	than	IN
zkie-zero-2020	193	10	both	both	DT
zkie-zero-2020	193	11	.	.	.
zkie-zero-2020	194	1	Data	Data	NNP
zkie-zero-2020	194	2	Model	Model	NNP
zkie-zero-2020	194	3	A@1	A@1	NNS
zkie-zero-2020	194	4	A@5	A@5	NNP
zkie-zero-2020	194	5	|C|	|C|	NNP
zkie-zero-2020	194	6	t	t	NN
zkie-zero-2020	194	7	AIDA	AIDA	NNP
zkie-zero-2020	194	8	MFLEB	MFLEB	NNP
zkie-zero-2020	194	9	0.56	0.56	CD
zkie-zero-2020	194	10	0.71	0.71	CD
zkie-zero-2020	194	11	31	31	CD
zkie-zero-2020	194	12	LightGBM	LightGBM	NNP
zkie-zero-2020	194	13	0.44	0.44	CD
zkie-zero-2020	194	14	0.72	0.72	CD
zkie-zero-2020	194	15	9	9	CD
zkie-zero-2020	194	16	RankSVM	ranksvm	CD
zkie-zero-2020	194	17	0.37	0.37	CD
zkie-zero-2020	194	18	0.69	0.69	CD
zkie-zero-2020	194	19	56	56	CD
zkie-zero-2020	194	20	RankNet	RankNet	NNP
zkie-zero-2020	194	21	0.42	0.42	CD
zkie-zero-2020	194	22	0.70	0.70	CD
zkie-zero-2020	194	23	190	190	CD
zkie-zero-2020	194	24	WWO	WWO	NNP
zkie-zero-2020	194	25	MFLEB	MFLEB	NNP
zkie-zero-2020	194	26	0.32	0.32	CD
zkie-zero-2020	194	27	0.77	0.77	CD
zkie-zero-2020	194	28	19	19	CD
zkie-zero-2020	194	29	LightGBM	LightGBM	NNS
zkie-zero-2020	194	30	0.37	0.37	CD
zkie-zero-2020	194	31	0.83	0.83	CD
zkie-zero-2020	194	32	2	2	CD
zkie-zero-2020	194	33	RankSVM	ranksvm	CD
zkie-zero-2020	194	34	0.46	0.46	CD
zkie-zero-2020	194	35	0.86	0.86	CD
zkie-zero-2020	194	36	15	15	CD
zkie-zero-2020	194	37	RankNet	RankNet	NNP
zkie-zero-2020	194	38	0.52	0.52	CD
zkie-zero-2020	194	39	0.87	0.87	CD
zkie-zero-2020	194	40	37	37	CD
zkie-zero-2020	194	41	1641	1641	CD
zkie-zero-2020	194	42	MFLEB	MFLEB	NNP
zkie-zero-2020	194	43	0.28	0.28	CD
zkie-zero-2020	194	44	0.75	0.75	CD
zkie-zero-2020	194	45	38	38	CD
zkie-zero-2020	194	46	LightGBM	LightGBM	NNP
zkie-zero-2020	194	47	0.35	0.35	CD
zkie-zero-2020	194	48	0.77	0.77	CD
zkie-zero-2020	194	49	1	1	CD
zkie-zero-2020	194	50	RankSVM	ranksvm	CD
zkie-zero-2020	194	51	0.48	0.48	CD
zkie-zero-2020	194	52	0.80	0.80	CD
zkie-zero-2020	194	53	1	1	CD
zkie-zero-2020	194	54	RankNet	RankNet	NNP
zkie-zero-2020	194	55	0.55	0.55	CD
zkie-zero-2020	194	56	0.83	0.83	CD
zkie-zero-2020	194	57	2	2	CD
zkie-zero-2020	194	58	Table	table	NN
zkie-zero-2020	194	59	4	4	CD
zkie-zero-2020	194	60	:	:	:
zkie-zero-2020	194	61	Ranking	rank	VBG
zkie-zero-2020	194	62	scores	score	NNS
zkie-zero-2020	194	63	when	when	WRB
zkie-zero-2020	194	64	using	use	VBG
zkie-zero-2020	194	65	all	all	PDT
zkie-zero-2020	194	66	the	the	DT
zkie-zero-2020	194	67	data	datum	NNS
zkie-zero-2020	194	68	.	.	.
zkie-zero-2020	195	1	We	-PRON-	PRP
zkie-zero-2020	195	2	report	report	VBP
zkie-zero-2020	195	3	Accuracy@1	Accuracy@1	.
zkie-zero-2020	195	4	(	(	-LRB-
zkie-zero-2020	195	5	Gold	Gold	NNP
zkie-zero-2020	195	6	Candidate	Candidate	NNP
zkie-zero-2020	195	7	was	be	VBD
zkie-zero-2020	195	8	ranked	rank	VBN
zkie-zero-2020	195	9	high-	high-	JJ
zkie-zero-2020	195	10	est	est	NNP
zkie-zero-2020	195	11	,	,	,
zkie-zero-2020	195	12	Accuracy@5	Accuracy@5	.
zkie-zero-2020	195	13	(	(	-LRB-
zkie-zero-2020	195	14	Gold	Gold	NNP
zkie-zero-2020	195	15	Candidate	Candidate	NNP
zkie-zero-2020	195	16	was	be	VBD
zkie-zero-2020	195	17	in	in	IN
zkie-zero-2020	195	18	top	top	JJ
zkie-zero-2020	195	19	5	5	CD
zkie-zero-2020	195	20	predic-	predic-	NN
zkie-zero-2020	195	21	tions	tion	NNS
zkie-zero-2020	195	22	of	of	IN
zkie-zero-2020	195	23	the	the	DT
zkie-zero-2020	195	24	ranker	ranker	NN
zkie-zero-2020	195	25	)	)	-RRB-
zkie-zero-2020	195	26	)	)	-RRB-
zkie-zero-2020	195	27	.	.	.
zkie-zero-2020	196	1	|C|	|C|	NNP
zkie-zero-2020	196	2	denotes	denote	VBZ
zkie-zero-2020	196	3	the	the	DT
zkie-zero-2020	196	4	average	average	JJ
zkie-zero-2020	196	5	number	number	NN
zkie-zero-2020	196	6	of	of	IN
zkie-zero-2020	196	7	candidates	candidate	NNS
zkie-zero-2020	196	8	found	find	VBN
zkie-zero-2020	196	9	for	for	IN
zkie-zero-2020	196	10	each	each	DT
zkie-zero-2020	196	11	mention	mention	NN
zkie-zero-2020	196	12	.	.	.
zkie-zero-2020	197	1	For	for	IN
zkie-zero-2020	197	2	AIDA	AIDA	NNP
zkie-zero-2020	197	3	,	,	,
zkie-zero-2020	197	4	we	-PRON-	PRP
zkie-zero-2020	197	5	evaluate	evaluate	VBP
zkie-zero-2020	197	6	on	on	IN
zkie-zero-2020	197	7	the	the	DT
zkie-zero-2020	197	8	test	test	NN
zkie-zero-2020	197	9	set	set	NN
zkie-zero-2020	197	10	,	,	,
zkie-zero-2020	197	11	for	for	IN
zkie-zero-2020	197	12	the	the	DT
zkie-zero-2020	197	13	other	other	JJ
zkie-zero-2020	197	14	datasets	dataset	NNS
zkie-zero-2020	197	15	,	,	,
zkie-zero-2020	197	16	we	-PRON-	PRP
zkie-zero-2020	197	17	use	use	VBP
zkie-zero-2020	197	18	10-fold	10-fold	CD
zkie-zero-2020	197	19	cross	cross	NN
zkie-zero-2020	197	20	validation	validation	NN
zkie-zero-2020	197	21	.	.	.
zkie-zero-2020	198	1	We	-PRON-	PRP
zkie-zero-2020	198	2	also	also	RB
zkie-zero-2020	198	3	measure	measure	VBP
zkie-zero-2020	198	4	the	the	DT
zkie-zero-2020	198	5	training	training	NN
zkie-zero-2020	198	6	time	time	NN
zkie-zero-2020	198	7	t	t	NN
zkie-zero-2020	198	8	in	in	IN
zkie-zero-2020	198	9	seconds	second	NNS
zkie-zero-2020	198	10	averaged	average	VBN
zkie-zero-2020	198	11	over	over	IN
zkie-zero-2020	198	12	10	10	CD
zkie-zero-2020	198	13	runs	run	NNS
zkie-zero-2020	198	14	.	.	.
zkie-zero-2020	199	1	Feature	feature	NN
zkie-zero-2020	199	2	importance	importance	NN
zkie-zero-2020	199	3	The	the	DT
zkie-zero-2020	199	4	models	model	NNS
zkie-zero-2020	199	5	we	-PRON-	PRP
zkie-zero-2020	199	6	chose	choose	VBD
zkie-zero-2020	199	7	for	for	IN
zkie-zero-2020	199	8	ranking	rank	VBG
zkie-zero-2020	199	9	are	be	VBP
zkie-zero-2020	199	10	white	white	JJ
zkie-zero-2020	199	11	-	-	HYPH
zkie-zero-2020	199	12	box	box	NN
zkie-zero-2020	199	13	;	;	:
zkie-zero-2020	199	14	they	-PRON-	PRP
zkie-zero-2020	199	15	allow	allow	VBP
zkie-zero-2020	199	16	us	-PRON-	PRP
zkie-zero-2020	199	17	to	to	TO
zkie-zero-2020	199	18	introspect	introspect	VB
zkie-zero-2020	199	19	the	the	DT
zkie-zero-2020	199	20	importance	importance	NN
zkie-zero-2020	199	21	they	-PRON-	PRP
zkie-zero-2020	199	22	give	give	VBP
zkie-zero-2020	199	23	to	to	IN
zkie-zero-2020	199	24	each	each	DT
zkie-zero-2020	199	25	feature	feature	NN
zkie-zero-2020	199	26	,	,	,
zkie-zero-2020	199	27	thereby	thereby	RB
zkie-zero-2020	199	28	explaining	explain	VBG
zkie-zero-2020	199	29	their	-PRON-	PRP$
zkie-zero-2020	199	30	scoring	score	VBG
zkie-zero-2020	199	31	choice	choice	NN
zkie-zero-2020	199	32	.	.	.
zkie-zero-2020	200	1	For	for	IN
zkie-zero-2020	200	2	the	the	DT
zkie-zero-2020	200	3	RankSVM	ranksvm	NN
zkie-zero-2020	200	4	,	,	,
zkie-zero-2020	200	5	we	-PRON-	PRP
zkie-zero-2020	200	6	follow	follow	VBP
zkie-zero-2020	200	7	Guyon	Guyon	NNP
zkie-zero-2020	200	8	et	et	NNP
zkie-zero-2020	200	9	al	al	NNP
zkie-zero-2020	200	10	.	.	.
zkie-zero-2020	201	1	(	(	-LRB-
zkie-zero-2020	201	2	2002	2002	CD
zkie-zero-2020	201	3	)	)	-RRB-
zkie-zero-2020	201	4	and	and	CC
zkie-zero-2020	201	5	use	use	VB
zkie-zero-2020	201	6	the	the	DT
zkie-zero-2020	201	7	square	square	NN
zkie-zero-2020	201	8	of	of	IN
zkie-zero-2020	201	9	the	the	DT
zkie-zero-2020	201	10	model	model	NN
zkie-zero-2020	201	11	weights	weight	VBZ
zkie-zero-2020	201	12	as	as	IN
zkie-zero-2020	201	13	importance	importance	NN
zkie-zero-2020	201	14	.	.	.
zkie-zero-2020	202	1	For	for	IN
zkie-zero-2020	202	2	Light-	Light-	NNP
zkie-zero-2020	202	3	GBM	GBM	NNP
zkie-zero-2020	202	4	,	,	,
zkie-zero-2020	202	5	we	-PRON-	PRP
zkie-zero-2020	202	6	use	use	VBP
zkie-zero-2020	202	7	the	the	DT
zkie-zero-2020	202	8	number	number	NN
zkie-zero-2020	202	9	of	of	IN
zkie-zero-2020	202	10	times	time	NNS
zkie-zero-2020	202	11	a	a	DT
zkie-zero-2020	202	12	feature	feature	NN
zkie-zero-2020	202	13	is	be	VBZ
zkie-zero-2020	202	14	used	use	VBN
zkie-zero-2020	202	15	to	to	TO
zkie-zero-2020	202	16	make	make	VB
zkie-zero-2020	202	17	a	a	DT
zkie-zero-2020	202	18	split	split	NN
zkie-zero-2020	202	19	in	in	IN
zkie-zero-2020	202	20	a	a	DT
zkie-zero-2020	202	21	decision	decision	NN
zkie-zero-2020	202	22	tree	tree	NN
zkie-zero-2020	202	23	.	.	.
zkie-zero-2020	203	1	We	-PRON-	PRP
zkie-zero-2020	203	2	train	train	VBP
zkie-zero-2020	203	3	RankSVM	RankSVM	NNP
zkie-zero-2020	203	4	and	and	CC
zkie-zero-2020	203	5	LightGBM	LightGBM	NNP
zkie-zero-2020	203	6	models	model	NNS
zkie-zero-2020	203	7	on	on	IN
zkie-zero-2020	203	8	all	all	DT
zkie-zero-2020	203	9	data	datum	NNS
zkie-zero-2020	203	10	and	and	CC
zkie-zero-2020	203	11	report	report	VB
zkie-zero-2020	203	12	the	the	DT
zkie-zero-2020	203	13	most	most	RBS
zkie-zero-2020	203	14	important	important	JJ
zkie-zero-2020	203	15	and	and	CC
zkie-zero-2020	203	16	least	least	RBS
zkie-zero-2020	203	17	important	important	JJ
zkie-zero-2020	203	18	fea-	fea-	JJ
zkie-zero-2020	203	19	tures	ture	NNS
zkie-zero-2020	203	20	in	in	IN
zkie-zero-2020	203	21	Fig	Fig	NNP
zkie-zero-2020	203	22	.	.	.
zkie-zero-2020	204	1	3	3	LS
zkie-zero-2020	204	2	.	.	.
zkie-zero-2020	205	1	We	-PRON-	PRP
zkie-zero-2020	205	2	normalize	normalize	VBP
zkie-zero-2020	205	3	the	the	DT
zkie-zero-2020	205	4	weights	weight	NNS
zkie-zero-2020	205	5	by	by	IN
zkie-zero-2020	205	6	the	the	DT
zkie-zero-2020	205	7	L1-norm	l1-norm	NN
zkie-zero-2020	205	8	.	.	.
zkie-zero-2020	206	1	It	-PRON-	PRP
zkie-zero-2020	206	2	can	can	MD
zkie-zero-2020	206	3	be	be	VB
zkie-zero-2020	206	4	seen	see	VBN
zkie-zero-2020	206	5	that	that	IN
zkie-zero-2020	206	6	both	both	DT
zkie-zero-2020	206	7	models	model	NNS
zkie-zero-2020	206	8	rely	rely	VBP
zkie-zero-2020	206	9	on	on	IN
zkie-zero-2020	206	10	Levenshtein	Levenshtein	NNP
zkie-zero-2020	206	11	distance	distance	NN
zkie-zero-2020	206	12	between	between	IN
zkie-zero-2020	206	13	mention	mention	NN
zkie-zero-2020	206	14	and	and	CC
zkie-zero-2020	206	15	label	label	NN
zkie-zero-2020	206	16	as	as	RB
zkie-zero-2020	206	17	well	well	RB
zkie-zero-2020	206	18	as	as	IN
zkie-zero-2020	206	19	Sentence	Sentence	NNP
zkie-zero-2020	206	20	-	-	HYPH
zkie-zero-2020	206	21	BERT	BERT	NNP
zkie-zero-2020	206	22	.	.	.
zkie-zero-2020	207	1	The	the	DT
zkie-zero-2020	207	2	other	other	JJ
zkie-zero-2020	207	3	text	text	NN
zkie-zero-2020	207	4	similarity	similarity	NN
zkie-zero-2020	207	5	features	feature	NNS
zkie-zero-2020	207	6	are	be	VBP
zkie-zero-2020	207	7	,	,	,
zkie-zero-2020	207	8	while	while	IN
zkie-zero-2020	207	9	sparingly	sparingly	RB
zkie-zero-2020	207	10	,	,	,
zkie-zero-2020	207	11	also	also	RB
zkie-zero-2020	207	12	used	use	VBD
zkie-zero-2020	207	13	.	.	.
zkie-zero-2020	208	1	Simple	simple	JJ
zkie-zero-2020	208	2	fea-	fea-	JJ
zkie-zero-2020	208	3	tures	ture	NNS
zkie-zero-2020	208	4	like	like	IN
zkie-zero-2020	208	5	exact	exact	JJ
zkie-zero-2020	208	6	match	match	NN
zkie-zero-2020	208	7	,	,	,
zkie-zero-2020	208	8	contains	contain	NNS
zkie-zero-2020	208	9	or	or	CC
zkie-zero-2020	208	10	prefix	prefix	NN
zkie-zero-2020	208	11	and	and	CC
zkie-zero-2020	208	12	postfix	postfix	VB
zkie-zero-2020	208	13	seem	seem	VBP
zkie-zero-2020	208	14	to	to	TO
zkie-zero-2020	208	15	not	not	RB
zkie-zero-2020	208	16	have	have	VB
zkie-zero-2020	208	17	a	a	DT
zkie-zero-2020	208	18	large	large	JJ
zkie-zero-2020	208	19	impact	impact	NN
zkie-zero-2020	208	20	.	.	.
zkie-zero-2020	209	1	In	in	IN
zkie-zero-2020	209	2	general	general	JJ
zkie-zero-2020	209	3	,	,	,
zkie-zero-2020	209	4	LightGBM	LightGBM	NNP
zkie-zero-2020	209	5	uses	use	VBZ
zkie-zero-2020	209	6	more	more	JJR
zkie-zero-2020	209	7	features	feature	NNS
zkie-zero-2020	209	8	than	than	IN
zkie-zero-2020	209	9	the	the	DT
zkie-zero-2020	209	10	RankSVM	ranksvm	NN
zkie-zero-2020	209	11	.	.	.
zkie-zero-2020	210	1	Even	even	RB
zkie-zero-2020	210	2	though	though	IN
zkie-zero-2020	210	3	Sentence	Sentence	NNP
zkie-zero-2020	210	4	-	-	HYPH
zkie-zero-2020	210	5	BERT	BERT	NNP
zkie-zero-2020	210	6	was	be	VBD
zkie-zero-2020	210	7	trained	train	VBN
zkie-zero-2020	210	8	on	on	IN
zkie-zero-2020	210	9	Natural	Natural	NNP
zkie-zero-2020	210	10	Language	Language	NNP
zkie-zero-2020	210	11	Inference	Inference	NNP
zkie-zero-2020	210	12	(	(	-LRB-
zkie-zero-2020	210	13	NLI	NLI	NNP
zkie-zero-2020	210	14	)	)	-RRB-
zkie-zero-2020	210	15	data	datum	NNS
zkie-zero-2020	210	16	which	which	WDT
zkie-zero-2020	210	17	contains	contain	VBZ
zkie-zero-2020	210	18	only	only	RB
zkie-zero-2020	210	19	relatively	relatively	RB
zkie-zero-2020	210	20	simple	simple	JJ
zkie-zero-2020	210	21	sentences	sentence	NNS
zkie-zero-2020	210	22	,	,	,
zkie-zero-2020	210	23	it	-PRON-	PRP
zkie-zero-2020	210	24	still	still	RB
zkie-zero-2020	210	25	is	be	VBZ
zkie-zero-2020	210	26	relied	rely	VBN
zkie-zero-2020	210	27	on	on	RP
zkie-zero-2020	210	28	by	by	IN
zkie-zero-2020	210	29	both	both	DT
zkie-zero-2020	210	30	models	model	NNS
zkie-zero-2020	210	31	for	for	IN
zkie-zero-2020	210	32	all	all	DT
zkie-zero-2020	210	33	datasets	dataset	NNS
zkie-zero-2020	210	34	.	.	.
zkie-zero-2020	211	1	The	the	DT
zkie-zero-2020	211	2	high	high	JJ
zkie-zero-2020	211	3	importance	importance	NN
zkie-zero-2020	211	4	of	of	IN
zkie-zero-2020	211	5	Levenshtein	Levenshtein	NNP
zkie-zero-2020	211	6	distance	distance	NN
zkie-zero-2020	211	7	between	between	IN
zkie-zero-2020	211	8	mention	mention	NN
zkie-zero-2020	211	9	and	and	CC
zkie-zero-2020	211	10	label	label	NN
zkie-zero-2020	211	11	for	for	IN
zkie-zero-2020	211	12	1641	1641	CD
zkie-zero-2020	211	13	is	be	VBZ
zkie-zero-2020	211	14	expected	expect	VBN
zkie-zero-2020	211	15	and	and	CC
zkie-zero-2020	211	16	can	can	MD
zkie-zero-2020	211	17	be	be	VB
zkie-zero-2020	211	18	explained	explain	VBN
zkie-zero-2020	211	19	by	by	IN
zkie-zero-2020	211	20	the	the	DT
zkie-zero-2020	211	21	fact	fact	NN
zkie-zero-2020	211	22	that	that	IN
zkie-zero-2020	211	23	the	the	DT
zkie-zero-2020	211	24	knowledge	knowledge	NN
zkie-zero-2020	211	25	base	base	NN
zkie-zero-2020	211	26	labels	label	NNS
zkie-zero-2020	211	27	often	often	RB
zkie-zero-2020	211	28	were	be	VBD
zkie-zero-2020	211	29	derived	derive	VBN
zkie-zero-2020	211	30	from	from	IN
zkie-zero-2020	211	31	the	the	DT
zkie-zero-2020	211	32	mentions	mention	NNS
zkie-zero-2020	211	33	in	in	IN
zkie-zero-2020	211	34	the	the	DT
zkie-zero-2020	211	35	text	text	NN
zkie-zero-2020	211	36	when	when	WRB
zkie-zero-2020	211	37	creating	create	VBG
zkie-zero-2020	211	38	a	a	DT
zkie-zero-2020	211	39	domain	domain	NN
zkie-zero-2020	211	40	-	-	HYPH
zkie-zero-2020	211	41	specific	specific	JJ
zkie-zero-2020	211	42	knowledge	knowledge	NN
zkie-zero-2020	211	43	6988	6988	CD
zkie-zero-2020	211	44	base	base	NN
zkie-zero-2020	211	45	for	for	IN
zkie-zero-2020	211	46	this	this	DT
zkie-zero-2020	211	47	dataset	dataset	NN
zkie-zero-2020	211	48	.	.	.
zkie-zero-2020	212	1	When	when	WRB
zkie-zero-2020	212	2	trained	train	VBN
zkie-zero-2020	212	3	on	on	IN
zkie-zero-2020	212	4	AIDA	AIDA	NNP
zkie-zero-2020	212	5	,	,	,
zkie-zero-2020	212	6	the	the	DT
zkie-zero-2020	212	7	RankSVM	RankSVM	NNP
zkie-zero-2020	212	8	assigns	assign	VBZ
zkie-zero-2020	212	9	a	a	DT
zkie-zero-2020	212	10	high	high	JJ
zkie-zero-2020	212	11	importance	importance	NN
zkie-zero-2020	212	12	to	to	IN
zkie-zero-2020	212	13	the	the	DT
zkie-zero-2020	212	14	Jac-	Jac-	NNP
zkie-zero-2020	212	15	card	card	NN
zkie-zero-2020	212	16	distance	distance	NN
zkie-zero-2020	212	17	between	between	IN
zkie-zero-2020	212	18	context	context	NN
zkie-zero-2020	212	19	and	and	CC
zkie-zero-2020	212	20	description	description	NN
zkie-zero-2020	212	21	.	.	.
zkie-zero-2020	213	1	We	-PRON-	PRP
zkie-zero-2020	213	2	attribute	attribute	VBP
zkie-zero-2020	213	3	this	this	DT
zkie-zero-2020	213	4	to	to	IN
zkie-zero-2020	213	5	the	the	DT
zkie-zero-2020	213	6	fact	fact	NN
zkie-zero-2020	213	7	that	that	IN
zkie-zero-2020	213	8	entity	entity	NN
zkie-zero-2020	213	9	descriptions	description	NNS
zkie-zero-2020	213	10	in	in	IN
zkie-zero-2020	213	11	Wikidata	Wikidata	NNP
zkie-zero-2020	213	12	are	be	VBP
zkie-zero-2020	213	13	quite	quite	RB
zkie-zero-2020	213	14	short	short	JJ
zkie-zero-2020	213	15	;	;	:
zkie-zero-2020	213	16	if	if	IN
zkie-zero-2020	213	17	they	-PRON-	PRP
zkie-zero-2020	213	18	are	be	VBP
zkie-zero-2020	213	19	similar	similar	JJ
zkie-zero-2020	213	20	to	to	IN
zkie-zero-2020	213	21	the	the	DT
zkie-zero-2020	213	22	context	context	NN
zkie-zero-2020	213	23	then	then	RB
zkie-zero-2020	213	24	it	-PRON-	PRP
zkie-zero-2020	213	25	is	be	VBZ
zkie-zero-2020	213	26	very	very	RB
zkie-zero-2020	213	27	likely	likely	RB
zkie-zero-2020	213	28	a	a	DT
zkie-zero-2020	213	29	match	match	NN
zkie-zero-2020	213	30	.	.	.
zkie-zero-2020	214	1	Figure	figure	NN
zkie-zero-2020	214	2	3	3	CD
zkie-zero-2020	214	3	:	:	:
zkie-zero-2020	214	4	Feature	feature	NN
zkie-zero-2020	214	5	importance	importance	NN
zkie-zero-2020	214	6	of	of	IN
zkie-zero-2020	214	7	the	the	DT
zkie-zero-2020	214	8	respective	respective	JJ
zkie-zero-2020	214	9	models	model	NNS
zkie-zero-2020	214	10	for	for	IN
zkie-zero-2020	214	11	different	different	JJ
zkie-zero-2020	214	12	datasets	dataset	NNS
zkie-zero-2020	214	13	.	.	.
zkie-zero-2020	215	1	For	for	IN
zkie-zero-2020	215	2	the	the	DT
zkie-zero-2020	215	3	RankSVM	ranksvm	NN
zkie-zero-2020	215	4	,	,	,
zkie-zero-2020	215	5	we	-PRON-	PRP
zkie-zero-2020	215	6	use	use	VBP
zkie-zero-2020	215	7	the	the	DT
zkie-zero-2020	215	8	squared	squared	JJ
zkie-zero-2020	215	9	weights	weight	NNS
zkie-zero-2020	215	10	;	;	,
zkie-zero-2020	215	11	for	for	IN
zkie-zero-2020	215	12	LightGBM	LightGBM	NNP
zkie-zero-2020	215	13	,	,	,
zkie-zero-2020	215	14	we	-PRON-	PRP
zkie-zero-2020	215	15	use	use	VBP
zkie-zero-2020	215	16	the	the	DT
zkie-zero-2020	215	17	number	number	NN
zkie-zero-2020	215	18	of	of	IN
zkie-zero-2020	215	19	times	time	NNS
zkie-zero-2020	215	20	a	a	DT
zkie-zero-2020	215	21	feature	feature	NN
zkie-zero-2020	215	22	is	be	VBZ
zkie-zero-2020	215	23	used	use	VBN
zkie-zero-2020	215	24	for	for	IN
zkie-zero-2020	215	25	splitting	splitting	NN
zkie-zero-2020	215	26	.	.	.
zkie-zero-2020	216	1	Both	both	DT
zkie-zero-2020	216	2	are	be	VBP
zkie-zero-2020	216	3	normal-	normal-	RB
zkie-zero-2020	216	4	ized	ize	VBN
zkie-zero-2020	216	5	to	to	TO
zkie-zero-2020	216	6	sum	sum	VB
zkie-zero-2020	216	7	up	up	RP
zkie-zero-2020	216	8	to	to	IN
zkie-zero-2020	216	9	1	1	CD
zkie-zero-2020	216	10	.	.	.
zkie-zero-2020	217	1	ML	ML	NNP
zkie-zero-2020	217	2	stands	stand	VBZ
zkie-zero-2020	217	3	for	for	IN
zkie-zero-2020	217	4	Mention	Mention	NNP
zkie-zero-2020	217	5	-	-	HYPH
zkie-zero-2020	217	6	Label	Label	NNP
zkie-zero-2020	217	7	,	,	,
zkie-zero-2020	217	8	CD	CD	NNP
zkie-zero-2020	217	9	for	for	IN
zkie-zero-2020	217	10	Context	Context	NNP
zkie-zero-2020	217	11	-	-	HYPH
zkie-zero-2020	217	12	Description	Description	NNP
zkie-zero-2020	217	13	.	.	.
zkie-zero-2020	218	1	5.3	5.3	CD
zkie-zero-2020	218	2	Simulation	simulation	NN
zkie-zero-2020	218	3	We	-PRON-	PRP
zkie-zero-2020	218	4	simulate	simulate	VBP
zkie-zero-2020	218	5	the	the	DT
zkie-zero-2020	218	6	Human	human	NN
zkie-zero-2020	218	7	-	-	HYPH
zkie-zero-2020	218	8	In	in	IN
zkie-zero-2020	218	9	-	-	HYPH
zkie-zero-2020	218	10	The	the	DT
zkie-zero-2020	218	11	-	-	HYPH
zkie-zero-2020	218	12	Loop	Loop	NNP
zkie-zero-2020	218	13	setting	setting	NN
zkie-zero-2020	218	14	by	by	IN
zkie-zero-2020	218	15	modeling	model	VBG
zkie-zero-2020	218	16	a	a	DT
zkie-zero-2020	218	17	user	user	NN
zkie-zero-2020	218	18	annotating	annotate	VBG
zkie-zero-2020	218	19	an	an	DT
zkie-zero-2020	218	20	unannotated	unannotated	JJ
zkie-zero-2020	218	21	corpus	corpus	NN
zkie-zero-2020	218	22	linearly	linearly	RB
zkie-zero-2020	218	23	.	.	.
zkie-zero-2020	219	1	In	in	IN
zkie-zero-2020	219	2	the	the	DT
zkie-zero-2020	219	3	beginning	beginning	NN
zkie-zero-2020	219	4	,	,	,
zkie-zero-2020	219	5	they	-PRON-	PRP
zkie-zero-2020	219	6	annotate	annotate	VBP
zkie-zero-2020	219	7	an	an	DT
zkie-zero-2020	219	8	ini-	ini-	NN
zkie-zero-2020	219	9	tial	tial	JJ
zkie-zero-2020	219	10	seed	seed	NN
zkie-zero-2020	219	11	of	of	IN
zkie-zero-2020	219	12	10	10	CD
zkie-zero-2020	219	13	entities	entity	NNS
zkie-zero-2020	219	14	without	without	IN
zkie-zero-2020	219	15	annotation	annotation	NN
zkie-zero-2020	219	16	support	support	NN
zkie-zero-2020	219	17	which	which	WDT
zkie-zero-2020	219	18	are	be	VBP
zkie-zero-2020	219	19	then	then	RB
zkie-zero-2020	219	20	used	use	VBN
zkie-zero-2020	219	21	to	to	TO
zkie-zero-2020	219	22	bootstrap	bootstrap	VB
zkie-zero-2020	219	23	the	the	DT
zkie-zero-2020	219	24	ranker	ranker	JJR
zkie-zero-2020	219	25	.	.	.
zkie-zero-2020	220	1	At	at	IN
zkie-zero-2020	220	2	every	every	DT
zkie-zero-2020	220	3	step	step	NN
zkie-zero-2020	220	4	,	,	,
zkie-zero-2020	220	5	the	the	DT
zkie-zero-2020	220	6	user	user	NN
zkie-zero-2020	220	7	annotates	annotate	VBZ
zkie-zero-2020	220	8	several	several	JJ
zkie-zero-2020	220	9	entities	entity	NNS
zkie-zero-2020	220	10	where	where	WRB
zkie-zero-2020	220	11	the	the	DT
zkie-zero-2020	220	12	ranker	ranker	NNP
zkie-zero-2020	220	13	is	be	VBZ
zkie-zero-2020	220	14	used	use	VBN
zkie-zero-2020	220	15	as	as	IN
zkie-zero-2020	220	16	assistance	assistance	NN
zkie-zero-2020	220	17	.	.	.
zkie-zero-2020	221	1	After	after	IN
zkie-zero-2020	221	2	an	an	DT
zkie-zero-2020	221	3	anno-	anno-	JJ
zkie-zero-2020	221	4	tation	tation	NN
zkie-zero-2020	221	5	batch	batch	NN
zkie-zero-2020	221	6	is	be	VBZ
zkie-zero-2020	221	7	finished	finish	VBN
zkie-zero-2020	221	8	,	,	,
zkie-zero-2020	221	9	this	this	DT
zkie-zero-2020	221	10	new	new	JJ
zkie-zero-2020	221	11	data	datum	NNS
zkie-zero-2020	221	12	is	be	VBZ
zkie-zero-2020	221	13	added	add	VBN
zkie-zero-2020	221	14	to	to	IN
zkie-zero-2020	221	15	the	the	DT
zkie-zero-2020	221	16	training	training	NN
zkie-zero-2020	221	17	set	set	NN
zkie-zero-2020	221	18	,	,	,
zkie-zero-2020	221	19	the	the	DT
zkie-zero-2020	221	20	ranker	ranker	JJR
zkie-zero-2020	221	21	is	be	VBZ
zkie-zero-2020	221	22	retrained	retrain	VBN
zkie-zero-2020	221	23	and	and	CC
zkie-zero-2020	221	24	evalu-	evalu-	NNS
zkie-zero-2020	221	25	ated	ate	VBN
zkie-zero-2020	221	26	.	.	.
zkie-zero-2020	222	1	Only	only	RB
zkie-zero-2020	222	2	LightGBM	LightGBM	NNP
zkie-zero-2020	222	3	and	and	CC
zkie-zero-2020	222	4	RankSVM	RankSVM	NNP
zkie-zero-2020	222	5	are	be	VBP
zkie-zero-2020	222	6	used	use	VBN
zkie-zero-2020	222	7	as	as	IN
zkie-zero-2020	222	8	the	the	DT
zkie-zero-2020	222	9	RankNet	RankNet	NNP
zkie-zero-2020	222	10	turned	turn	VBD
zkie-zero-2020	222	11	out	out	RP
zkie-zero-2020	222	12	to	to	TO
zkie-zero-2020	222	13	be	be	VB
zkie-zero-2020	222	14	too	too	RB
zkie-zero-2020	222	15	slow	slow	JJ
zkie-zero-2020	222	16	.	.	.
zkie-zero-2020	223	1	We	-PRON-	PRP
zkie-zero-2020	223	2	do	do	VBP
zkie-zero-2020	223	3	not	not	RB
zkie-zero-2020	223	4	evaluate	evaluate	VB
zkie-zero-2020	223	5	on	on	IN
zkie-zero-2020	223	6	a	a	DT
zkie-zero-2020	223	7	holdout	holdout	NN
zkie-zero-2020	223	8	set	set	NN
zkie-zero-2020	223	9	.	.	.
zkie-zero-2020	224	1	Instead	instead	RB
zkie-zero-2020	224	2	,	,	,
zkie-zero-2020	224	3	we	-PRON-	PRP
zkie-zero-2020	224	4	follow	follow	VBP
zkie-zero-2020	224	5	Erdmann	Erdmann	NNP
zkie-zero-2020	224	6	et	et	FW
zkie-zero-2020	224	7	al	al	NNP
zkie-zero-2020	224	8	.	.	.
zkie-zero-2020	225	1	(	(	-LRB-
zkie-zero-2020	225	2	2019	2019	CD
zkie-zero-2020	225	3	)	)	-RRB-
zkie-zero-2020	225	4	and	and	CC
zkie-zero-2020	225	5	simulate	simulate	VB
zkie-zero-2020	225	6	annotating	annotate	VBG
zkie-zero-2020	225	7	the	the	DT
zkie-zero-2020	225	8	complete	complete	JJ
zkie-zero-2020	225	9	corpus	corpus	NN
zkie-zero-2020	225	10	and	and	CC
zkie-zero-2020	225	11	evaluate	evaluate	VB
zkie-zero-2020	225	12	on	on	IN
zkie-zero-2020	225	13	the	the	DT
zkie-zero-2020	225	14	very	very	RB
zkie-zero-2020	225	15	same	same	JJ
zkie-zero-2020	225	16	data	datum	NNS
zkie-zero-2020	225	17	as	as	IN
zkie-zero-2020	225	18	we	-PRON-	PRP
zkie-zero-2020	225	19	are	be	VBP
zkie-zero-2020	225	20	interested	interested	JJ
zkie-zero-2020	225	21	in	in	IN
zkie-zero-2020	225	22	how	how	WRB
zkie-zero-2020	225	23	an	an	DT
zkie-zero-2020	225	24	annotated	annotate	VBN
zkie-zero-2020	225	25	sub-	sub-	NN
zkie-zero-2020	225	26	set	set	NN
zkie-zero-2020	225	27	helps	help	VBZ
zkie-zero-2020	225	28	to	to	TO
zkie-zero-2020	225	29	annotate	annotate	VB
zkie-zero-2020	225	30	the	the	DT
zkie-zero-2020	225	31	rest	rest	NN
zkie-zero-2020	225	32	of	of	IN
zkie-zero-2020	225	33	the	the	DT
zkie-zero-2020	225	34	data	datum	NNS
zkie-zero-2020	225	35	,	,	,
zkie-zero-2020	225	36	not	not	RB
zkie-zero-2020	225	37	how	how	WRB
zkie-zero-2020	225	38	well	well	RB
zkie-zero-2020	225	39	the	the	DT
zkie-zero-2020	225	40	model	model	NN
zkie-zero-2020	225	41	generalizes	generalize	VBZ
zkie-zero-2020	225	42	.	.	.
zkie-zero-2020	226	1	We	-PRON-	PRP
zkie-zero-2020	226	2	assume	assume	VBP
zkie-zero-2020	226	3	that	that	IN
zkie-zero-2020	226	4	users	user	NNS
zkie-zero-2020	226	5	annotate	annotate	VBP
zkie-zero-2020	226	6	mention	mention	NN
zkie-zero-2020	226	7	spans	span	NNS
zkie-zero-2020	226	8	perfectly	perfectly	RB
zkie-zero-2020	226	9	,	,	,
zkie-zero-2020	226	10	i.e.	i.e.	FW
zkie-zero-2020	227	1	we	-PRON-	PRP
zkie-zero-2020	227	2	use	use	VBP
zkie-zero-2020	227	3	gold	gold	NN
zkie-zero-2020	227	4	spans	span	NNS
zkie-zero-2020	227	5	.	.	.
zkie-zero-2020	228	1	The	the	DT
zkie-zero-2020	228	2	candidate	candidate	NN
zkie-zero-2020	228	3	generation	generation	NN
zkie-zero-2020	228	4	is	be	VBZ
zkie-zero-2020	228	5	simulated	simulate	VBN
zkie-zero-2020	228	6	in	in	IN
zkie-zero-2020	228	7	three	three	CD
zkie-zero-2020	228	8	phases	phase	NNS
zkie-zero-2020	228	9	.	.	.
zkie-zero-2020	229	1	It	-PRON-	PRP
zkie-zero-2020	229	2	relies	rely	VBZ
zkie-zero-2020	229	3	on	on	IN
zkie-zero-2020	229	4	the	the	DT
zkie-zero-2020	229	5	fact	fact	NN
zkie-zero-2020	229	6	that	that	IN
zkie-zero-2020	229	7	the	the	DT
zkie-zero-2020	229	8	gold	gold	NN
zkie-zero-2020	229	9	en-	en-	XX
zkie-zero-2020	229	10	tity	tity	NN
zkie-zero-2020	229	11	is	be	VBZ
zkie-zero-2020	229	12	given	give	VBN
zkie-zero-2020	229	13	by	by	IN
zkie-zero-2020	229	14	the	the	DT
zkie-zero-2020	229	15	dataset	dataset	NN
zkie-zero-2020	229	16	:	:	:
zkie-zero-2020	229	17	First	first	RB
zkie-zero-2020	229	18	,	,	,
zkie-zero-2020	229	19	search	search	VB
zkie-zero-2020	229	20	for	for	IN
zkie-zero-2020	229	21	the	the	DT
zkie-zero-2020	229	22	mention	mention	NN
zkie-zero-2020	229	23	only	only	RB
zkie-zero-2020	229	24	.	.	.
zkie-zero-2020	230	1	If	if	IN
zkie-zero-2020	230	2	it	-PRON-	PRP
zkie-zero-2020	230	3	was	be	VBD
zkie-zero-2020	230	4	not	not	RB
zkie-zero-2020	230	5	found	find	VBN
zkie-zero-2020	230	6	,	,	,
zkie-zero-2020	230	7	search	search	VB
zkie-zero-2020	230	8	for	for	IN
zkie-zero-2020	230	9	the	the	DT
zkie-zero-2020	230	10	first	first	JJ
zkie-zero-2020	230	11	word	word	NN
zkie-zero-2020	230	12	of	of	IN
zkie-zero-2020	230	13	the	the	DT
zkie-zero-2020	230	14	mention	mention	NN
zkie-zero-2020	230	15	only	only	RB
zkie-zero-2020	230	16	.	.	.
zkie-zero-2020	231	1	If	if	IN
zkie-zero-2020	231	2	this	this	DT
zkie-zero-2020	231	3	does	do	VBZ
zkie-zero-2020	231	4	not	not	RB
zkie-zero-2020	231	5	return	return	VB
zkie-zero-2020	231	6	the	the	DT
zkie-zero-2020	231	7	gold	gold	JJ
zkie-zero-2020	231	8	entity	entity	NN
zkie-zero-2020	231	9	,	,	,
zkie-zero-2020	231	10	search	search	VB
zkie-zero-2020	231	11	for	for	IN
zkie-zero-2020	231	12	the	the	DT
zkie-zero-2020	231	13	gold	gold	JJ
zkie-zero-2020	231	14	entity	entity	NN
zkie-zero-2020	231	15	label	label	NN
zkie-zero-2020	231	16	.	.	.
zkie-zero-2020	232	1	All	all	DT
zkie-zero-2020	232	2	candidates	candidate	NNS
zkie-zero-2020	232	3	retrieved	retrieve	VBN
zkie-zero-2020	232	4	by	by	IN
zkie-zero-2020	232	5	these	these	DT
zkie-zero-2020	232	6	searches	search	NNS
zkie-zero-2020	232	7	for	for	IN
zkie-zero-2020	232	8	a	a	DT
zkie-zero-2020	232	9	mention	mention	NN
zkie-zero-2020	232	10	are	be	VBP
zkie-zero-2020	232	11	used	use	VBN
zkie-zero-2020	232	12	as	as	IN
zkie-zero-2020	232	13	training	training	NN
zkie-zero-2020	232	14	data	datum	NNS
zkie-zero-2020	232	15	.	.	.
zkie-zero-2020	233	1	We	-PRON-	PRP
zkie-zero-2020	233	2	also	also	RB
zkie-zero-2020	233	3	experimented	experiment	VBD
zkie-zero-2020	233	4	with	with	IN
zkie-zero-2020	233	5	using	use	VBG
zkie-zero-2020	233	6	only	only	JJ
zkie-zero-2020	233	7	candidates	candidate	NNS
zkie-zero-2020	233	8	for	for	IN
zkie-zero-2020	233	9	that	that	DT
zkie-zero-2020	233	10	the	the	DT
zkie-zero-2020	233	11	ranker	ranker	RBR
zkie-zero-2020	233	12	assigned	assign	VBD
zkie-zero-2020	233	13	a	a	DT
zkie-zero-2020	233	14	higher	high	JJR
zkie-zero-2020	233	15	score	score	NN
zkie-zero-2020	233	16	than	than	IN
zkie-zero-2020	233	17	the	the	DT
zkie-zero-2020	233	18	gold	gold	NN
zkie-zero-2020	233	19	one	one	CD
zkie-zero-2020	233	20	.	.	.
zkie-zero-2020	234	1	This	this	DT
zkie-zero-2020	234	2	,	,	,
zkie-zero-2020	234	3	however	however	RB
zkie-zero-2020	234	4	,	,	,
zkie-zero-2020	234	5	did	do	VBD
zkie-zero-2020	234	6	not	not	RB
zkie-zero-2020	234	7	affect	affect	VB
zkie-zero-2020	234	8	the	the	DT
zkie-zero-2020	234	9	performance	performance	NN
zkie-zero-2020	234	10	.	.	.
zkie-zero-2020	235	1	Therefore	therefore	RB
zkie-zero-2020	235	2	,	,	,
zkie-zero-2020	235	3	we	-PRON-	PRP
zkie-zero-2020	235	4	use	use	VBP
zkie-zero-2020	235	5	all	all	DT
zkie-zero-2020	235	6	negative	negative	JJ
zkie-zero-2020	235	7	candidates	candidate	NNS
zkie-zero-2020	235	8	.	.	.
zkie-zero-2020	236	1	Fig	fig	NN
zkie-zero-2020	236	2	.	.	.
zkie-zero-2020	237	1	4	4	CD
zkie-zero-2020	237	2	depicts	depict	VBZ
zkie-zero-2020	237	3	the	the	DT
zkie-zero-2020	237	4	simulation	simulation	NN
zkie-zero-2020	237	5	results	result	NNS
zkie-zero-2020	237	6	.	.	.
zkie-zero-2020	238	1	All	all	DT
zkie-zero-2020	238	2	mod-	mod-	NN
zkie-zero-2020	238	3	els	els	XX
zkie-zero-2020	238	4	outperform	outperform	JJ
zkie-zero-2020	238	5	the	the	DT
zkie-zero-2020	238	6	MFLE	MFLE	NNP
zkie-zero-2020	238	7	baseline	baseline	VBP
zkie-zero-2020	238	8	over	over	IN
zkie-zero-2020	238	9	most	most	JJS
zkie-zero-2020	238	10	of	of	IN
zkie-zero-2020	238	11	the	the	DT
zkie-zero-2020	238	12	annotation	annotation	NN
zkie-zero-2020	238	13	process	process	NN
zkie-zero-2020	238	14	.	.	.
zkie-zero-2020	239	1	It	-PRON-	PRP
zkie-zero-2020	239	2	can	can	MD
zkie-zero-2020	239	3	be	be	VB
zkie-zero-2020	239	4	seen	see	VBN
zkie-zero-2020	239	5	that	that	IN
zkie-zero-2020	239	6	both	both	DT
zkie-zero-2020	239	7	of	of	IN
zkie-zero-2020	239	8	our	-PRON-	PRP$
zkie-zero-2020	239	9	used	use	VBN
zkie-zero-2020	239	10	models	model	NNS
zkie-zero-2020	239	11	achieve	achieve	VBP
zkie-zero-2020	239	12	high	high	JJ
zkie-zero-2020	239	13	performance	performance	NN
zkie-zero-2020	239	14	even	even	RB
zkie-zero-2020	239	15	if	if	IN
zkie-zero-2020	239	16	trained	train	VBN
zkie-zero-2020	239	17	on	on	IN
zkie-zero-2020	239	18	very	very	RB
zkie-zero-2020	239	19	few	few	JJ
zkie-zero-2020	239	20	annotations	annotation	NNS
zkie-zero-2020	239	21	.	.	.
zkie-zero-2020	240	1	The	the	DT
zkie-zero-2020	240	2	RankSVM	RankSVM	``
zkie-zero-2020	240	3	handles	handle	VBZ
zkie-zero-2020	240	4	low	low	JJ
zkie-zero-2020	240	5	data	datum	NNS
zkie-zero-2020	240	6	better	well	JJR
zkie-zero-2020	240	7	than	than	IN
zkie-zero-2020	240	8	LightGBM	LightGBM	NNP
zkie-zero-2020	240	9	,	,	,
zkie-zero-2020	240	10	but	but	CC
zkie-zero-2020	240	11	quickly	quickly	RB
zkie-zero-2020	240	12	reaches	reach	VBZ
zkie-zero-2020	240	13	its	-PRON-	PRP$
zkie-zero-2020	240	14	peak	peak	NN
zkie-zero-2020	240	15	performance	performance	NN
zkie-zero-2020	240	16	due	due	IN
zkie-zero-2020	240	17	to	to	IN
zkie-zero-2020	240	18	it	-PRON-	PRP
zkie-zero-2020	240	19	be-	be-	XX
zkie-zero-2020	240	20	ing	e	VBG
zkie-zero-2020	240	21	a	a	DT
zkie-zero-2020	240	22	linear	linear	JJ
zkie-zero-2020	240	23	model	model	NN
zkie-zero-2020	240	24	with	with	IN
zkie-zero-2020	240	25	limited	limited	JJ
zkie-zero-2020	240	26	learning	learning	NN
zkie-zero-2020	240	27	capacity	capacity	NN
zkie-zero-2020	240	28	.	.	.
zkie-zero-2020	241	1	The	the	DT
zkie-zero-2020	241	2	LightGBM	LightGBM	NNP
zkie-zero-2020	241	3	does	do	VBZ
zkie-zero-2020	241	4	not	not	RB
zkie-zero-2020	241	5	plateau	plateau	VB
zkie-zero-2020	241	6	that	that	RB
zkie-zero-2020	241	7	early	early	RB
zkie-zero-2020	241	8	.	.	.
zkie-zero-2020	242	1	This	this	DT
zkie-zero-2020	242	2	potentially	potentially	RB
zkie-zero-2020	242	3	allows	allow	VBZ
zkie-zero-2020	242	4	to	to	TO
zkie-zero-2020	242	5	first	first	RB
zkie-zero-2020	242	6	use	use	VB
zkie-zero-2020	242	7	a	a	DT
zkie-zero-2020	242	8	RankSVM	ranksvm	NN
zkie-zero-2020	242	9	for	for	IN
zkie-zero-2020	242	10	the	the	DT
zkie-zero-2020	242	11	cold	cold	JJ
zkie-zero-2020	242	12	start	start	NN
zkie-zero-2020	242	13	and	and	CC
zkie-zero-2020	242	14	when	when	WRB
zkie-zero-2020	242	15	enough	enough	JJ
zkie-zero-2020	242	16	annotations	annotation	NNS
zkie-zero-2020	242	17	are	be	VBP
zkie-zero-2020	242	18	made	make	VBN
zkie-zero-2020	242	19	,	,	,
zkie-zero-2020	242	20	LightGBM	LightGBM	NNP
zkie-zero-2020	242	21	,	,	,
zkie-zero-2020	242	22	thereby	thereby	RB
zkie-zero-2020	242	23	combining	combine	VBG
zkie-zero-2020	242	24	the	the	DT
zkie-zero-2020	242	25	best	good	JJS
zkie-zero-2020	242	26	of	of	IN
zkie-zero-2020	242	27	both	both	DT
zkie-zero-2020	242	28	models	model	NNS
zkie-zero-2020	242	29	.	.	.
zkie-zero-2020	243	1	Comparing	compare	VBG
zkie-zero-2020	243	2	the	the	DT
zkie-zero-2020	243	3	performance	performance	NN
zkie-zero-2020	243	4	on	on	IN
zkie-zero-2020	243	5	the	the	DT
zkie-zero-2020	243	6	three	three	CD
zkie-zero-2020	243	7	datasets	dataset	NNS
zkie-zero-2020	243	8	,	,	,
zkie-zero-2020	243	9	we	-PRON-	PRP
zkie-zero-2020	243	10	notice	notice	VBP
zkie-zero-2020	243	11	that	that	IN
zkie-zero-2020	243	12	the	the	DT
zkie-zero-2020	243	13	performance	performance	NN
zkie-zero-2020	243	14	for	for	IN
zkie-zero-2020	243	15	AIDA	AIDA	NNP
zkie-zero-2020	243	16	is	be	VBZ
zkie-zero-2020	243	17	much	much	RB
zkie-zero-2020	243	18	higher	high	JJR
zkie-zero-2020	243	19	.	.	.
zkie-zero-2020	244	1	Also	also	RB
zkie-zero-2020	244	2	,	,	,
zkie-zero-2020	244	3	the	the	DT
zkie-zero-2020	244	4	baseline	baseline	NN
zkie-zero-2020	244	5	rises	rise	VBZ
zkie-zero-2020	244	6	much	much	RB
zkie-zero-2020	244	7	more	more	RBR
zkie-zero-2020	244	8	steeply	steeply	RB
zkie-zero-2020	244	9	,	,	,
zkie-zero-2020	244	10	hinting	hint	VBG
zkie-zero-2020	244	11	again	again	RB
zkie-zero-2020	244	12	that	that	IN
zkie-zero-2020	244	13	AIDA	AIDA	NNP
zkie-zero-2020	244	14	is	be	VBZ
zkie-zero-2020	244	15	easier	easy	JJR
zkie-zero-2020	244	16	and	and	CC
zkie-zero-2020	244	17	pop-	pop-	NN
zkie-zero-2020	244	18	ularity	ularity	NN
zkie-zero-2020	244	19	there	there	EX
zkie-zero-2020	244	20	is	be	VBZ
zkie-zero-2020	244	21	a	a	DT
zkie-zero-2020	244	22	very	very	RB
zkie-zero-2020	244	23	strong	strong	JJ
zkie-zero-2020	244	24	feature	feature	NN
zkie-zero-2020	244	25	.	.	.
zkie-zero-2020	245	1	For	for	IN
zkie-zero-2020	245	2	1641	1641	CD
zkie-zero-2020	245	3	,	,	,
zkie-zero-2020	245	4	the	the	DT
zkie-zero-2020	245	5	curve	curve	NN
zkie-zero-2020	245	6	continue	continue	VBP
zkie-zero-2020	245	7	to	to	TO
zkie-zero-2020	245	8	rise	rise	VB
zkie-zero-2020	245	9	,	,	,
zkie-zero-2020	245	10	hinting	hint	VBG
zkie-zero-2020	245	11	that	that	IN
zkie-zero-2020	245	12	more	more	JJR
zkie-zero-2020	245	13	data	datum	NNS
zkie-zero-2020	245	14	is	be	VBZ
zkie-zero-2020	245	15	needed	need	VBN
zkie-zero-2020	245	16	to	to	TO
zkie-zero-2020	245	17	reach	reach	VB
zkie-zero-2020	245	18	maximum	maximum	JJ
zkie-zero-2020	245	19	performance	performance	NN
zkie-zero-2020	245	20	.	.	.
zkie-zero-2020	246	1	Dataset	Dataset	NNP
zkie-zero-2020	246	2	Phase	Phase	NNP
zkie-zero-2020	246	3	1	1	CD
zkie-zero-2020	246	4	Phase	phase	NN
zkie-zero-2020	246	5	2	2	CD
zkie-zero-2020	246	6	Phase	phase	NN
zkie-zero-2020	246	7	3	3	CD
zkie-zero-2020	246	8	AIDA	aida	NN
zkie-zero-2020	246	9	0.20	0.20	CD
zkie-zero-2020	246	10	0.00	0.00	CD
zkie-zero-2020	246	11	0.80	0.80	CD
zkie-zero-2020	246	12	WWO	WWO	NNP
zkie-zero-2020	246	13	0.26	0.26	CD
zkie-zero-2020	246	14	0.27	0.27	CD
zkie-zero-2020	246	15	0.47	0.47	CD
zkie-zero-2020	246	16	1641	1641	CD
zkie-zero-2020	246	17	0.55	0.55	CD
zkie-zero-2020	246	18	0.06	0.06	CD
zkie-zero-2020	246	19	0.39	0.39	CD
zkie-zero-2020	246	20	Table	table	NN
zkie-zero-2020	246	21	5	5	CD
zkie-zero-2020	246	22	:	:	:
zkie-zero-2020	246	23	Percentage	percentage	NN
zkie-zero-2020	246	24	of	of	IN
zkie-zero-2020	246	25	times	time	NNS
zkie-zero-2020	246	26	the	the	DT
zkie-zero-2020	246	27	simulated	simulate	VBN
zkie-zero-2020	246	28	user	user	NN
zkie-zero-2020	246	29	found	find	VBD
zkie-zero-2020	246	30	the	the	DT
zkie-zero-2020	246	31	gold	gold	JJ
zkie-zero-2020	246	32	entity	entity	NN
zkie-zero-2020	246	33	in	in	IN
zkie-zero-2020	246	34	the	the	DT
zkie-zero-2020	246	35	candidate	candidate	NN
zkie-zero-2020	246	36	list	list	NN
zkie-zero-2020	246	37	by	by	IN
zkie-zero-2020	246	38	searching	search	VBG
zkie-zero-2020	246	39	for	for	IN
zkie-zero-2020	246	40	the	the	DT
zkie-zero-2020	246	41	mention	mention	NN
zkie-zero-2020	246	42	(	(	-LRB-
zkie-zero-2020	246	43	Phase	phase	NN
zkie-zero-2020	246	44	1	1	CD
zkie-zero-2020	246	45	)	)	-RRB-
zkie-zero-2020	246	46	,	,	,
zkie-zero-2020	246	47	for	for	IN
zkie-zero-2020	246	48	the	the	DT
zkie-zero-2020	246	49	first	first	JJ
zkie-zero-2020	246	50	word	word	NN
zkie-zero-2020	246	51	of	of	IN
zkie-zero-2020	246	52	the	the	DT
zkie-zero-2020	246	53	mention	mention	NN
zkie-zero-2020	246	54	(	(	-LRB-
zkie-zero-2020	246	55	Phase	phase	NN
zkie-zero-2020	246	56	2	2	CD
zkie-zero-2020	246	57	)	)	-RRB-
zkie-zero-2020	246	58	or	or	CC
zkie-zero-2020	246	59	for	for	IN
zkie-zero-2020	246	60	the	the	DT
zkie-zero-2020	246	61	gold	gold	JJ
zkie-zero-2020	246	62	label	label	NN
zkie-zero-2020	246	63	(	(	-LRB-
zkie-zero-2020	246	64	Phase	phase	NN
zkie-zero-2020	246	65	3	3	CD
zkie-zero-2020	246	66	)	)	-RRB-
zkie-zero-2020	246	67	.	.	.
zkie-zero-2020	247	1	Table	table	NN
zkie-zero-2020	247	2	5	5	CD
zkie-zero-2020	247	3	shows	show	VBZ
zkie-zero-2020	247	4	how	how	WRB
zkie-zero-2020	247	5	the	the	DT
zkie-zero-2020	247	6	simulated	simulated	JJ
zkie-zero-2020	247	7	user	user	NN
zkie-zero-2020	247	8	searched	search	VBD
zkie-zero-2020	247	9	for	for	IN
zkie-zero-2020	247	10	the	the	DT
zkie-zero-2020	247	11	gold	gold	NN
zkie-zero-2020	247	12	entities	entity	NNS
zkie-zero-2020	247	13	.	.	.
zkie-zero-2020	248	1	We	-PRON-	PRP
zkie-zero-2020	248	2	see	see	VBP
zkie-zero-2020	248	3	that	that	IN
zkie-zero-2020	248	4	for	for	IN
zkie-zero-2020	248	5	WWO	WWO	NNP
zkie-zero-2020	248	6	and	and	CC
zkie-zero-2020	248	7	1641	1641	CD
zkie-zero-2020	248	8	,	,	,
zkie-zero-2020	248	9	the	the	DT
zkie-zero-2020	248	10	user	user	NN
zkie-zero-2020	248	11	often	often	RB
zkie-zero-2020	248	12	does	do	VBZ
zkie-zero-2020	248	13	not	not	RB
zkie-zero-2020	248	14	need	need	VB
zkie-zero-2020	248	15	to	to	TO
zkie-zero-2020	248	16	spend	spend	VB
zkie-zero-2020	248	17	much	much	JJ
zkie-zero-2020	248	18	effort	effort	NN
zkie-zero-2020	248	19	in	in	IN
zkie-zero-2020	248	20	searching	search	VBG
zkie-zero-2020	248	21	for	for	IN
zkie-zero-2020	248	22	the	the	DT
zkie-zero-2020	248	23	gold	gold	JJ
zkie-zero-2020	248	24	label	label	NN
zkie-zero-2020	248	25	,	,	,
zkie-zero-2020	248	26	using	use	VBG
zkie-zero-2020	248	27	the	the	DT
zkie-zero-2020	248	28	mention	mention	NN
zkie-zero-2020	248	29	is	be	VBZ
zkie-zero-2020	248	30	in	in	IN
zkie-zero-2020	248	31	around	around	RB
zkie-zero-2020	248	32	50	50	CD
zkie-zero-2020	248	33	%	%	NN
zkie-zero-2020	248	34	of	of	IN
zkie-zero-2020	248	35	the	the	DT
zkie-zero-2020	248	36	cases	case	NNS
zkie-zero-2020	248	37	enough	enough	RB
zkie-zero-2020	248	38	.	.	.
zkie-zero-2020	249	1	We	-PRON-	PRP
zkie-zero-2020	249	2	attribute	attribute	VBP
zkie-zero-2020	249	3	this	this	DT
zkie-zero-2020	249	4	to	to	IN
zkie-zero-2020	249	5	the	the	DT
zkie-zero-2020	249	6	fuzzy	fuzzy	JJ
zkie-zero-2020	249	7	search	search	NN
zkie-zero-2020	249	8	which	which	WDT
zkie-zero-2020	249	9	the	the	DT
zkie-zero-2020	249	10	official	official	JJ
zkie-zero-2020	249	11	Wikidata	Wikidata	NNP
zkie-zero-2020	249	12	endpoint	endpoint	NN
zkie-zero-2020	249	13	does	do	VBZ
zkie-zero-2020	249	14	not	not	RB
zkie-zero-2020	249	15	offer	offer	VB
zkie-zero-2020	249	16	.	.	.
zkie-zero-2020	250	1	6989	6989	CD
zkie-zero-2020	250	2	0	0	CD
zkie-zero-2020	250	3	5k	5k	CD
zkie-zero-2020	250	4	10k	10k	NNS
zkie-zero-2020	250	5	15k	15k	NNS
zkie-zero-2020	250	6	20k	20k	CD
zkie-zero-2020	250	7	25k	25k	NNS
zkie-zero-2020	250	8	30k	30k	CD
zkie-zero-2020	250	9	35k	35k	CD
zkie-zero-2020	250	10	0.4	0.4	CD
zkie-zero-2020	250	11	0.5	0.5	CD
zkie-zero-2020	250	12	0.6	0.6	CD
zkie-zero-2020	250	13	0.7	0.7	CD
zkie-zero-2020	250	14	0.8	0.8	CD
zkie-zero-2020	250	15	0.9	0.9	CD
zkie-zero-2020	250	16	1.0	1.0	CD
zkie-zero-2020	250	17	A	a	DT
zkie-zero-2020	250	18	cc	cc	NN
zkie-zero-2020	250	19	ur	ur	NN
zkie-zero-2020	250	20	ac	ac	IN
zkie-zero-2020	250	21	y@	y@	NNP
zkie-zero-2020	250	22	1	1	CD
zkie-zero-2020	250	23	0	0	CD
zkie-zero-2020	250	24	5k	5k	CD
zkie-zero-2020	250	25	10k	10k	NNS
zkie-zero-2020	250	26	15k	15k	NNS
zkie-zero-2020	250	27	20k	20k	CD
zkie-zero-2020	250	28	25k	25k	NNS
zkie-zero-2020	250	29	30k	30k	CD
zkie-zero-2020	250	30	35k	35k	NNS
zkie-zero-2020	250	31	Number	Number	NNP
zkie-zero-2020	250	32	of	of	IN
zkie-zero-2020	250	33	annotations	annotation	NNS
zkie-zero-2020	250	34	0.4	0.4	CD
zkie-zero-2020	250	35	0.5	0.5	CD
zkie-zero-2020	250	36	0.6	0.6	CD
zkie-zero-2020	250	37	0.7	0.7	CD
zkie-zero-2020	250	38	0.8	0.8	CD
zkie-zero-2020	250	39	0.9	0.9	CD
zkie-zero-2020	250	40	1.0	1.0	CD
zkie-zero-2020	250	41	A	a	DT
zkie-zero-2020	250	42	cc	cc	NN
zkie-zero-2020	250	43	ur	ur	NN
zkie-zero-2020	250	44	ac	ac	IN
zkie-zero-2020	250	45	y@	y@	NNP
zkie-zero-2020	250	46	5	5	CD
zkie-zero-2020	250	47	MFLE	mfle	NN
zkie-zero-2020	250	48	baseline	baseline	NN
zkie-zero-2020	250	49	LightGBM	LightGBM	NNP
zkie-zero-2020	250	50	RankSVM	RankSVM	VBZ
zkie-zero-2020	250	51	AIDA	AIDA	NNP
zkie-zero-2020	250	52	-	-	HYPH
zkie-zero-2020	250	53	CoNLL	CoNLL	NNP
zkie-zero-2020	250	54	0	0	CD
zkie-zero-2020	250	55	2.5k	2.5k	CD
zkie-zero-2020	250	56	5k	5k	CD
zkie-zero-2020	250	57	7.5k	7.5k	CD
zkie-zero-2020	250	58	10k	10k	CD
zkie-zero-2020	250	59	12.5k	12.5k	CD
zkie-zero-2020	250	60	0.4	0.4	CD
zkie-zero-2020	250	61	0.5	0.5	CD
zkie-zero-2020	250	62	0.6	0.6	CD
zkie-zero-2020	250	63	0.7	0.7	CD
zkie-zero-2020	250	64	0.8	0.8	CD
zkie-zero-2020	250	65	0.9	0.9	CD
zkie-zero-2020	250	66	1.0	1.0	CD
zkie-zero-2020	250	67	A	a	DT
zkie-zero-2020	250	68	cc	cc	NN
zkie-zero-2020	250	69	ur	ur	NN
zkie-zero-2020	250	70	ac	ac	IN
zkie-zero-2020	250	71	y@	y@	NNP
zkie-zero-2020	250	72	1	1	CD
zkie-zero-2020	250	73	0	0	CD
zkie-zero-2020	250	74	2.5k	2.5k	CD
zkie-zero-2020	250	75	5k	5k	CD
zkie-zero-2020	250	76	7.5k	7.5k	CD
zkie-zero-2020	250	77	10k	10k	CD
zkie-zero-2020	250	78	12.5k	12.5k	CD
zkie-zero-2020	250	79	Number	Number	NNP
zkie-zero-2020	250	80	of	of	IN
zkie-zero-2020	250	81	annotations	annotation	NNS
zkie-zero-2020	250	82	0.4	0.4	CD
zkie-zero-2020	250	83	0.5	0.5	CD
zkie-zero-2020	250	84	0.6	0.6	CD
zkie-zero-2020	250	85	0.7	0.7	CD
zkie-zero-2020	250	86	0.8	0.8	CD
zkie-zero-2020	250	87	0.9	0.9	CD
zkie-zero-2020	250	88	1.0	1.0	CD
zkie-zero-2020	250	89	A	a	DT
zkie-zero-2020	250	90	cc	cc	NN
zkie-zero-2020	250	91	ur	ur	NN
zkie-zero-2020	250	92	ac	ac	IN
zkie-zero-2020	250	93	y@	y@	NNP
zkie-zero-2020	250	94	5	5	CD
zkie-zero-2020	250	95	MFLE	mfle	NN
zkie-zero-2020	250	96	baseline	baseline	NN
zkie-zero-2020	250	97	LightGBM	LightGBM	.
zkie-zero-2020	250	98	RankSVM	RankSVM	.
zkie-zero-2020	250	99	Women	Women	NNP
zkie-zero-2020	250	100	Writers	Writers	NNP
zkie-zero-2020	250	101	0	0	CD
zkie-zero-2020	250	102	100	100	CD
zkie-zero-2020	250	103	200	200	CD
zkie-zero-2020	250	104	300	300	CD
zkie-zero-2020	250	105	400	400	CD
zkie-zero-2020	250	106	500	500	CD
zkie-zero-2020	250	107	0.4	0.4	CD
zkie-zero-2020	250	108	0.5	0.5	CD
zkie-zero-2020	250	109	0.6	0.6	CD
zkie-zero-2020	250	110	0.7	0.7	CD
zkie-zero-2020	250	111	0.8	0.8	CD
zkie-zero-2020	250	112	0.9	0.9	CD
zkie-zero-2020	250	113	1.0	1.0	CD
zkie-zero-2020	250	114	A	a	DT
zkie-zero-2020	250	115	cc	cc	NN
zkie-zero-2020	250	116	ur	ur	NN
zkie-zero-2020	250	117	ac	ac	IN
zkie-zero-2020	250	118	y@	y@	NNP
zkie-zero-2020	250	119	1	1	CD
zkie-zero-2020	250	120	0	0	CD
zkie-zero-2020	250	121	100	100	CD
zkie-zero-2020	250	122	200	200	CD
zkie-zero-2020	250	123	300	300	CD
zkie-zero-2020	250	124	400	400	CD
zkie-zero-2020	250	125	500	500	CD
zkie-zero-2020	250	126	Number	Number	NNP
zkie-zero-2020	250	127	of	of	IN
zkie-zero-2020	250	128	annotations	annotation	NNS
zkie-zero-2020	250	129	0.4	0.4	CD
zkie-zero-2020	250	130	0.5	0.5	CD
zkie-zero-2020	250	131	0.6	0.6	CD
zkie-zero-2020	250	132	0.7	0.7	CD
zkie-zero-2020	250	133	0.8	0.8	CD
zkie-zero-2020	250	134	0.9	0.9	CD
zkie-zero-2020	250	135	1.0	1.0	CD
zkie-zero-2020	250	136	A	a	DT
zkie-zero-2020	250	137	cc	cc	NN
zkie-zero-2020	250	138	ur	ur	NN
zkie-zero-2020	250	139	ac	ac	IN
zkie-zero-2020	250	140	y@	y@	NNP
zkie-zero-2020	250	141	5	5	CD
zkie-zero-2020	250	142	MFLE	mfle	NN
zkie-zero-2020	250	143	baseline	baseline	NN
zkie-zero-2020	250	144	LightGBM	LightGBM	.
zkie-zero-2020	250	145	RankSVM	RankSVM	NNP
zkie-zero-2020	250	146	1641	1641	CD
zkie-zero-2020	250	147	Depositions	Depositions	NNPS
zkie-zero-2020	250	148	Figure	Figure	NNP
zkie-zero-2020	250	149	4	4	CD
zkie-zero-2020	250	150	:	:	:
zkie-zero-2020	250	151	Human	human	NN
zkie-zero-2020	250	152	-	-	HYPH
zkie-zero-2020	250	153	in	in	IN
zkie-zero-2020	250	154	-	-	HYPH
zkie-zero-2020	250	155	the	the	DT
zkie-zero-2020	250	156	-	-	HYPH
zkie-zero-2020	250	157	loop	loop	NN
zkie-zero-2020	250	158	simulation	simulation	NN
zkie-zero-2020	250	159	results	result	NNS
zkie-zero-2020	250	160	for	for	IN
zkie-zero-2020	250	161	our	-PRON-	PRP$
zkie-zero-2020	250	162	three	three	CD
zkie-zero-2020	250	163	datasets	dataset	NNS
zkie-zero-2020	250	164	and	and	CC
zkie-zero-2020	250	165	models	model	NNS
zkie-zero-2020	250	166	.	.	.
zkie-zero-2020	251	1	We	-PRON-	PRP
zkie-zero-2020	251	2	can	can	MD
zkie-zero-2020	251	3	see	see	VB
zkie-zero-2020	251	4	that	that	IN
zkie-zero-2020	251	5	we	-PRON-	PRP
zkie-zero-2020	251	6	get	get	VBP
zkie-zero-2020	251	7	good	good	JJ
zkie-zero-2020	251	8	Accuracy@5	Accuracy@5	:
zkie-zero-2020	251	9	with	with	IN
zkie-zero-2020	251	10	only	only	RB
zkie-zero-2020	251	11	a	a	DT
zkie-zero-2020	251	12	few	few	JJ
zkie-zero-2020	251	13	annotations	annotation	NNS
zkie-zero-2020	251	14	,	,	,
zkie-zero-2020	251	15	especially	especially	RB
zkie-zero-2020	251	16	for	for	IN
zkie-zero-2020	251	17	the	the	DT
zkie-zero-2020	251	18	RankSVM	ranksvm	NN
zkie-zero-2020	251	19	.	.	.
zkie-zero-2020	252	1	This	this	DT
zkie-zero-2020	252	2	shows	show	VBZ
zkie-zero-2020	252	3	that	that	IN
zkie-zero-2020	252	4	the	the	DT
zkie-zero-2020	252	5	system	system	NN
zkie-zero-2020	252	6	is	be	VBZ
zkie-zero-2020	252	7	useful	useful	JJ
zkie-zero-2020	252	8	even	even	RB
zkie-zero-2020	252	9	at	at	IN
zkie-zero-2020	252	10	the	the	DT
zkie-zero-2020	252	11	beginning	beginning	NN
zkie-zero-2020	252	12	of	of	IN
zkie-zero-2020	252	13	the	the	DT
zkie-zero-2020	252	14	annotation	annotation	NN
zkie-zero-2020	252	15	process	process	NN
zkie-zero-2020	252	16	,	,	,
zkie-zero-2020	252	17	alleviating	alleviate	VBG
zkie-zero-2020	252	18	the	the	DT
zkie-zero-2020	252	19	cold	cold	JJ
zkie-zero-2020	252	20	start	start	NN
zkie-zero-2020	252	21	problem	problem	NN
zkie-zero-2020	252	22	.	.	.
zkie-zero-2020	253	1	5.4	5.4	CD
zkie-zero-2020	253	2	User	User	NNP
zkie-zero-2020	253	3	Study	Study	NNP
zkie-zero-2020	253	4	In	in	IN
zkie-zero-2020	253	5	order	order	NN
zkie-zero-2020	253	6	to	to	TO
zkie-zero-2020	253	7	validate	validate	VB
zkie-zero-2020	253	8	the	the	DT
zkie-zero-2020	253	9	viability	viability	NN
zkie-zero-2020	253	10	of	of	IN
zkie-zero-2020	253	11	our	-PRON-	PRP$
zkie-zero-2020	253	12	approach	approach	NN
zkie-zero-2020	253	13	in	in	IN
zkie-zero-2020	253	14	a	a	DT
zkie-zero-2020	253	15	realistic	realistic	JJ
zkie-zero-2020	253	16	scenario	scenario	NN
zkie-zero-2020	253	17	,	,	,
zkie-zero-2020	253	18	we	-PRON-	PRP
zkie-zero-2020	253	19	conduct	conduct	VBP
zkie-zero-2020	253	20	a	a	DT
zkie-zero-2020	253	21	user	user	NN
zkie-zero-2020	253	22	study	study	NN
zkie-zero-2020	253	23	.	.	.
zkie-zero-2020	254	1	For	for	IN
zkie-zero-2020	254	2	that	that	DT
zkie-zero-2020	254	3	,	,	,
zkie-zero-2020	254	4	we	-PRON-	PRP
zkie-zero-2020	254	5	augmented	augment	VBD
zkie-zero-2020	254	6	the	the	DT
zkie-zero-2020	254	7	already	already	RB
zkie-zero-2020	254	8	existing	exist	VBG
zkie-zero-2020	254	9	anno-	anno-	JJ
zkie-zero-2020	254	10	tation	tation	NN
zkie-zero-2020	254	11	tool	tool	NN
zkie-zero-2020	254	12	INCEpTION5	inception5	NN
zkie-zero-2020	254	13	(	(	-LRB-
zkie-zero-2020	254	14	Klie	Klie	NNP
zkie-zero-2020	254	15	et	et	FW
zkie-zero-2020	254	16	al	al	NNP
zkie-zero-2020	254	17	.	.	NNP
zkie-zero-2020	254	18	,	,	,
zkie-zero-2020	254	19	2018	2018	CD
zkie-zero-2020	254	20	)	)	-RRB-
zkie-zero-2020	254	21	with	with	IN
zkie-zero-2020	254	22	our	-PRON-	PRP$
zkie-zero-2020	254	23	Human	human	JJ
zkie-zero-2020	254	24	-	-	HYPH
zkie-zero-2020	254	25	In	in	IN
zkie-zero-2020	254	26	-	-	HYPH
zkie-zero-2020	254	27	The	the	DT
zkie-zero-2020	254	28	-	-	HYPH
zkie-zero-2020	254	29	Loop	Loop	NNP
zkie-zero-2020	254	30	entity	entity	NN
zkie-zero-2020	254	31	ranking	ranking	NN
zkie-zero-2020	254	32	and	and	CC
zkie-zero-2020	254	33	auto-	auto-	PRP$
zkie-zero-2020	254	34	matic	matic	JJ
zkie-zero-2020	254	35	suggestions	suggestion	NNS
zkie-zero-2020	254	36	.	.	.
zkie-zero-2020	255	1	Fig	fig	NN
zkie-zero-2020	255	2	.	.	.
zkie-zero-2020	256	1	5	5	CD
zkie-zero-2020	256	2	shows	show	VBZ
zkie-zero-2020	256	3	a	a	DT
zkie-zero-2020	256	4	screenshot	screenshot	NN
zkie-zero-2020	256	5	of	of	IN
zkie-zero-2020	256	6	the	the	DT
zkie-zero-2020	256	7	annotation	annotation	NN
zkie-zero-2020	256	8	editor	editor	NN
zkie-zero-2020	256	9	itself	-PRON-	PRP
zkie-zero-2020	256	10	.	.	.
zkie-zero-2020	257	1	We	-PRON-	PRP
zkie-zero-2020	257	2	let	let	VBP
zkie-zero-2020	257	3	five	five	CD
zkie-zero-2020	257	4	users	user	NNS
zkie-zero-2020	257	5	reanno-	reanno-	VBG
zkie-zero-2020	257	6	tate	tate	NN
zkie-zero-2020	257	7	parts	part	NNS
zkie-zero-2020	257	8	of	of	IN
zkie-zero-2020	257	9	the	the	DT
zkie-zero-2020	257	10	1641	1641	CD
zkie-zero-2020	257	11	corpus	corpus	NNP
zkie-zero-2020	257	12	.	.	.
zkie-zero-2020	258	1	It	-PRON-	PRP
zkie-zero-2020	258	2	was	be	VBD
zkie-zero-2020	258	3	chosen	choose	VBN
zkie-zero-2020	258	4	as	as	IN
zkie-zero-2020	258	5	it	-PRON-	PRP
zkie-zero-2020	258	6	has	have	VBZ
zkie-zero-2020	258	7	a	a	DT
zkie-zero-2020	258	8	high	high	JJ
zkie-zero-2020	258	9	density	density	NN
zkie-zero-2020	258	10	of	of	IN
zkie-zero-2020	258	11	entity	entity	NN
zkie-zero-2020	258	12	mentions	mention	NNS
zkie-zero-2020	258	13	while	while	IN
zkie-zero-2020	258	14	being	be	VBG
zkie-zero-2020	258	15	small	small	JJ
zkie-zero-2020	258	16	enough	enough	RB
zkie-zero-2020	258	17	to	to	TO
zkie-zero-2020	258	18	be	be	VB
zkie-zero-2020	258	19	annotated	annotate	VBN
zkie-zero-2020	258	20	in	in	RP
zkie-zero-2020	258	21	under	under	IN
zkie-zero-2020	258	22	one	one	CD
zkie-zero-2020	258	23	hour	hour	NN
zkie-zero-2020	258	24	.	.	.
zkie-zero-2020	259	1	Users	user	NNS
zkie-zero-2020	259	2	stem	stem	VBP
zkie-zero-2020	259	3	from	from	IN
zkie-zero-2020	259	4	various	various	JJ
zkie-zero-2020	259	5	academic	academic	JJ
zkie-zero-2020	259	6	backgrounds	background	NNS
zkie-zero-2020	259	7	,	,	,
zkie-zero-2020	259	8	e.g.	e.g.	RB
zkie-zero-2020	260	1	natural	natural	JJ
zkie-zero-2020	260	2	language	language	NN
zkie-zero-2020	260	3	processing	processing	NN
zkie-zero-2020	260	4	,	,	,
zkie-zero-2020	260	5	computer	computer	NN
zkie-zero-2020	260	6	science	science	NN
zkie-zero-2020	260	7	and	and	CC
zkie-zero-2020	260	8	digital	digital	JJ
zkie-zero-2020	260	9	humanities	humanity	NNS
zkie-zero-2020	260	10	.	.	.
zkie-zero-2020	261	1	Roughly	roughly	RB
zkie-zero-2020	261	2	half	half	NN
zkie-zero-2020	261	3	of	of	IN
zkie-zero-2020	261	4	them	-PRON-	PRP
zkie-zero-2020	261	5	have	have	VBP
zkie-zero-2020	261	6	previous	previous	JJ
zkie-zero-2020	261	7	experience	experience	NN
zkie-zero-2020	261	8	with	with	IN
zkie-zero-2020	261	9	annotating	annotating	NN
zkie-zero-2020	261	10	.	.	.
zkie-zero-2020	262	1	We	-PRON-	PRP
zkie-zero-2020	262	2	compare	compare	VBP
zkie-zero-2020	262	3	two	two	CD
zkie-zero-2020	262	4	configurations	configuration	NNS
zkie-zero-2020	262	5	:	:	:
zkie-zero-2020	262	6	one	one	PRP
zkie-zero-2020	262	7	uses	use	VBZ
zkie-zero-2020	262	8	our	-PRON-	PRP$
zkie-zero-2020	262	9	ranking	ranking	NN
zkie-zero-2020	262	10	and	and	CC
zkie-zero-2020	262	11	Lev-	Lev-	NNP
zkie-zero-2020	262	12	enshtein	enshtein	NNP
zkie-zero-2020	262	13	recommender	recommender	NNP
zkie-zero-2020	262	14	,	,	,
zkie-zero-2020	262	15	one	one	NN
zkie-zero-2020	262	16	uses	use	VBZ
zkie-zero-2020	262	17	the	the	DT
zkie-zero-2020	262	18	ranking	ranking	NN
zkie-zero-2020	262	19	of	of	IN
zkie-zero-2020	262	20	the	the	DT
zkie-zero-2020	262	21	full	full	JJ
zkie-zero-2020	262	22	text	text	NN
zkie-zero-2020	262	23	search	search	NN
zkie-zero-2020	262	24	with	with	IN
zkie-zero-2020	262	25	the	the	DT
zkie-zero-2020	262	26	string	string	NN
zkie-zero-2020	262	27	matching	match	VBG
zkie-zero-2020	262	28	recom-	recom-	NN
zkie-zero-2020	262	29	mender	mender	NN
zkie-zero-2020	262	30	.	.	.
zkie-zero-2020	263	1	We	-PRON-	PRP
zkie-zero-2020	263	2	randomly	randomly	RB
zkie-zero-2020	263	3	selected	select	VBD
zkie-zero-2020	263	4	eight	eight	CD
zkie-zero-2020	263	5	documents	document	NNS
zkie-zero-2020	263	6	which	which	WDT
zkie-zero-2020	263	7	we	-PRON-	PRP
zkie-zero-2020	263	8	split	split	VBD
zkie-zero-2020	263	9	in	in	IN
zkie-zero-2020	263	10	two	two	CD
zkie-zero-2020	263	11	sets	set	NNS
zkie-zero-2020	263	12	of	of	IN
zkie-zero-2020	263	13	four	four	CD
zkie-zero-2020	263	14	documents	document	NNS
zkie-zero-2020	263	15	.	.	.
zkie-zero-2020	264	1	To	to	TO
zkie-zero-2020	264	2	reduce	reduce	VB
zkie-zero-2020	264	3	bias	bias	NN
zkie-zero-2020	264	4	,	,	,
zkie-zero-2020	264	5	we	-PRON-	PRP
zkie-zero-2020	264	6	assign	assign	VBP
zkie-zero-2020	264	7	users	user	NNS
zkie-zero-2020	264	8	in	in	IN
zkie-zero-2020	264	9	four	four	CD
zkie-zero-2020	264	10	groups	group	NNS
zkie-zero-2020	264	11	based	base	VBN
zkie-zero-2020	264	12	on	on	IN
zkie-zero-2020	264	13	which	which	WDT
zkie-zero-2020	264	14	part	part	NN
zkie-zero-2020	264	15	and	and	CC
zkie-zero-2020	264	16	which	which	WDT
zkie-zero-2020	264	17	ranking	rank	VBG
zkie-zero-2020	264	18	they	-PRON-	PRP
zkie-zero-2020	264	19	use	use	VBP
zkie-zero-2020	264	20	first	first	RB
zkie-zero-2020	264	21	.	.	.
zkie-zero-2020	265	1	Users	user	NNS
zkie-zero-2020	265	2	are	be	VBP
zkie-zero-2020	265	3	given	give	VBN
zkie-zero-2020	265	4	detailed	detailed	JJ
zkie-zero-2020	265	5	instructions	instruction	NNS
zkie-zero-2020	265	6	and	and	CC
zkie-zero-2020	265	7	a	a	DT
zkie-zero-2020	265	8	warm-	warm-	NN
zkie-zero-2020	265	9	up	up	IN
zkie-zero-2020	265	10	document	document	NN
zkie-zero-2020	265	11	that	that	WDT
zkie-zero-2020	265	12	is	be	VBZ
zkie-zero-2020	265	13	not	not	RB
zkie-zero-2020	265	14	used	use	VBN
zkie-zero-2020	265	15	in	in	IN
zkie-zero-2020	265	16	the	the	DT
zkie-zero-2020	265	17	evaluation	evaluation	NN
zkie-zero-2020	265	18	to	to	TO
zkie-zero-2020	265	19	get	get	VB
zkie-zero-2020	265	20	used	use	VBN
zkie-zero-2020	265	21	to	to	IN
zkie-zero-2020	265	22	the	the	DT
zkie-zero-2020	265	23	annotation	annotation	NN
zkie-zero-2020	265	24	process	process	NN
zkie-zero-2020	265	25	.	.	.
zkie-zero-2020	266	1	We	-PRON-	PRP
zkie-zero-2020	266	2	measure	measure	VBP
zkie-zero-2020	266	3	annotation	annotation	NN
zkie-zero-2020	266	4	time	time	NN
zkie-zero-2020	266	5	,	,	,
zkie-zero-2020	266	6	number	number	NN
zkie-zero-2020	266	7	of	of	IN
zkie-zero-2020	266	8	suggestions	suggestion	NNS
zkie-zero-2020	266	9	used	use	VBN
zkie-zero-2020	266	10	and	and	CC
zkie-zero-2020	266	11	search	search	NN
zkie-zero-2020	266	12	queries	query	NNS
zkie-zero-2020	266	13	performed	perform	VBD
zkie-zero-2020	266	14	.	.	.
zkie-zero-2020	267	1	After	after	IN
zkie-zero-2020	267	2	the	the	DT
zkie-zero-2020	267	3	annotation	annotation	NN
zkie-zero-2020	267	4	is	be	VBZ
zkie-zero-2020	267	5	finished	finish	VBN
zkie-zero-2020	267	6	,	,	,
zkie-zero-2020	267	7	we	-PRON-	PRP
zkie-zero-2020	267	8	ask	ask	VBP
zkie-zero-2020	267	9	users	user	NNS
zkie-zero-2020	267	10	to	to	TO
zkie-zero-2020	267	11	fill	fill	VB
zkie-zero-2020	267	12	out	out	RP
zkie-zero-2020	267	13	a	a	DT
zkie-zero-2020	267	14	survey	survey	NN
zkie-zero-2020	267	15	asking	ask	VBG
zkie-zero-2020	267	16	which	which	WDT
zkie-zero-2020	267	17	system	system	NN
zkie-zero-2020	267	18	they	-PRON-	PRP
zkie-zero-2020	267	19	prefer	prefer	VBP
zkie-zero-2020	267	20	,	,	,
zkie-zero-2020	267	21	how	how	WRB
zkie-zero-2020	267	22	they	-PRON-	PRP
zkie-zero-2020	267	23	experienced	experience	VBD
zkie-zero-2020	267	24	the	the	DT
zkie-zero-2020	267	25	annotation	annotation	NN
zkie-zero-2020	267	26	process	process	NN
zkie-zero-2020	267	27	and	and	CC
zkie-zero-2020	267	28	what	what	WP
zkie-zero-2020	267	29	suggestions	suggestion	NNS
zkie-zero-2020	267	30	they	-PRON-	PRP
zkie-zero-2020	267	31	have	have	VBP
zkie-zero-2020	267	32	to	to	TO
zkie-zero-2020	267	33	improve	improve	VB
zkie-zero-2020	267	34	it	-PRON-	PRP
zkie-zero-2020	267	35	.	.	.
zkie-zero-2020	268	1	The	the	DT
zkie-zero-2020	268	2	evaluation	evaluation	NN
zkie-zero-2020	268	3	of	of	IN
zkie-zero-2020	268	4	the	the	DT
zkie-zero-2020	268	5	user	user	NN
zkie-zero-2020	268	6	study	study	NN
zkie-zero-2020	268	7	5	5	CD
zkie-zero-2020	268	8	https://inception-project.github.io	https://inception-project.github.io	NNP
zkie-zero-2020	268	9	shows	show	VBZ
zkie-zero-2020	268	10	that	that	IN
zkie-zero-2020	268	11	using	use	VBG
zkie-zero-2020	268	12	our	-PRON-	PRP$
zkie-zero-2020	268	13	approach	approach	NN
zkie-zero-2020	268	14	,	,	,
zkie-zero-2020	268	15	users	user	NNS
zkie-zero-2020	268	16	on	on	IN
zkie-zero-2020	268	17	average	average	JJ
zkie-zero-2020	268	18	annotated	annotate	VBN
zkie-zero-2020	268	19	35	35	CD
zkie-zero-2020	268	20	%	%	NN
zkie-zero-2020	268	21	faster	fast	RBR
zkie-zero-2020	268	22	and	and	CC
zkie-zero-2020	268	23	needed	need	VBD
zkie-zero-2020	268	24	15	15	CD
zkie-zero-2020	268	25	%	%	NN
zkie-zero-2020	268	26	less	less	JJR
zkie-zero-2020	268	27	search	search	NN
zkie-zero-2020	268	28	queries	query	NNS
zkie-zero-2020	268	29	.	.	.
zkie-zero-2020	269	1	Users	user	NNS
zkie-zero-2020	269	2	positively	positively	RB
zkie-zero-2020	269	3	commented	comment	VBD
zkie-zero-2020	269	4	on	on	IN
zkie-zero-2020	269	5	the	the	DT
zkie-zero-2020	269	6	rank-	rank-	NN
zkie-zero-2020	269	7	ing	ing	NN
zkie-zero-2020	269	8	performance	performance	NN
zkie-zero-2020	269	9	and	and	CC
zkie-zero-2020	269	10	the	the	DT
zkie-zero-2020	269	11	annotation	annotation	NN
zkie-zero-2020	269	12	suggestions	suggestion	NNS
zkie-zero-2020	269	13	for	for	IN
zkie-zero-2020	269	14	both	both	DT
zkie-zero-2020	269	15	systems	system	NNS
zkie-zero-2020	269	16	.	.	.
zkie-zero-2020	270	1	For	for	IN
zkie-zero-2020	270	2	our	-PRON-	PRP$
zkie-zero-2020	270	3	ranking	ranking	NN
zkie-zero-2020	270	4	,	,	,
zkie-zero-2020	270	5	users	user	NNS
zkie-zero-2020	270	6	reported	report	VBD
zkie-zero-2020	270	7	that	that	IN
zkie-zero-2020	270	8	the	the	DT
zkie-zero-2020	270	9	gold	gold	JJ
zkie-zero-2020	270	10	entity	entity	NN
zkie-zero-2020	270	11	often	often	RB
zkie-zero-2020	270	12	ranked	rank	VBD
zkie-zero-2020	270	13	first	first	RB
zkie-zero-2020	270	14	or	or	CC
zkie-zero-2020	270	15	close	close	RB
zkie-zero-2020	270	16	to	to	IN
zkie-zero-2020	270	17	top	top	NN
zkie-zero-2020	270	18	;	;	:
zkie-zero-2020	270	19	they	-PRON-	PRP
zkie-zero-2020	270	20	rarely	rarely	RB
zkie-zero-2020	270	21	observed	observe	VBD
zkie-zero-2020	270	22	that	that	IN
zkie-zero-2020	270	23	gold	gold	JJ
zkie-zero-2020	270	24	candidates	candidate	NNS
zkie-zero-2020	270	25	were	be	VBD
zkie-zero-2020	270	26	sorted	sort	VBN
zkie-zero-2020	270	27	close	close	RB
zkie-zero-2020	270	28	to	to	IN
zkie-zero-2020	270	29	the	the	DT
zkie-zero-2020	270	30	end	end	NN
zkie-zero-2020	270	31	of	of	IN
zkie-zero-2020	270	32	the	the	DT
zkie-zero-2020	270	33	candidate	candidate	NN
zkie-zero-2020	270	34	list	list	NN
zkie-zero-2020	270	35	.	.	.
zkie-zero-2020	271	1	We	-PRON-	PRP
zkie-zero-2020	271	2	conduct	conduct	VBP
zkie-zero-2020	271	3	a	a	DT
zkie-zero-2020	271	4	paired	paired	JJ
zkie-zero-2020	271	5	sample	sample	NN
zkie-zero-2020	271	6	t	t	NN
zkie-zero-2020	271	7	-	-	HYPH
zkie-zero-2020	271	8	test	test	NN
zkie-zero-2020	271	9	to	to	TO
zkie-zero-2020	271	10	estimate	estimate	VB
zkie-zero-2020	271	11	the	the	DT
zkie-zero-2020	271	12	significance	significance	NN
zkie-zero-2020	271	13	of	of	IN
zkie-zero-2020	271	14	our	-PRON-	PRP$
zkie-zero-2020	271	15	user	user	NN
zkie-zero-2020	271	16	study	study	NN
zkie-zero-2020	271	17	.	.	.
zkie-zero-2020	272	1	Our	-PRON-	PRP$
zkie-zero-2020	272	2	null	null	NN
zkie-zero-2020	272	3	-	-	HYPH
zkie-zero-2020	272	4	hypothesis	hypothesis	NN
zkie-zero-2020	272	5	is	be	VBZ
zkie-zero-2020	272	6	that	that	IN
zkie-zero-2020	272	7	the	the	DT
zkie-zero-2020	272	8	reranking	reranking	NN
zkie-zero-2020	272	9	system	system	NN
zkie-zero-2020	272	10	does	do	VBZ
zkie-zero-2020	272	11	not	not	RB
zkie-zero-2020	272	12	improve	improve	VB
zkie-zero-2020	272	13	the	the	DT
zkie-zero-2020	272	14	average	average	JJ
zkie-zero-2020	272	15	annotation	annotation	NN
zkie-zero-2020	272	16	time	time	NN
zkie-zero-2020	272	17	.	.	.
zkie-zero-2020	273	1	Conducting	conduct	VBG
zkie-zero-2020	273	2	the	the	DT
zkie-zero-2020	273	3	test	test	NN
zkie-zero-2020	273	4	yields	yield	VBZ
zkie-zero-2020	273	5	the	the	DT
zkie-zero-2020	273	6	following	follow	VBG
zkie-zero-2020	273	7	:	:	:
zkie-zero-2020	273	8	t	t	NN
zkie-zero-2020	273	9	=	=	SYM
zkie-zero-2020	273	10	3.332	3.332	CD
zkie-zero-2020	273	11	,	,	,
zkie-zero-2020	273	12	p	p	NN
zkie-zero-2020	273	13	=	=	SYM
zkie-zero-2020	273	14	0.029	0.029	CD
zkie-zero-2020	273	15	.	.	.
zkie-zero-2020	274	1	We	-PRON-	PRP
zkie-zero-2020	274	2	therefore	therefore	RB
zkie-zero-2020	274	3	reject	reject	VBP
zkie-zero-2020	274	4	the	the	DT
zkie-zero-2020	274	5	null	null	NN
zkie-zero-2020	274	6	hypothesis	hypothesis	NN
zkie-zero-2020	274	7	with	with	IN
zkie-zero-2020	274	8	p	p	NN
zkie-zero-2020	274	9	=	=	SYM
zkie-zero-2020	274	10	0.029	0.029	CD
zkie-zero-2020	274	11	<	<	XX
zkie-zero-2020	274	12	0.05	0.05	CD
zkie-zero-2020	274	13	,	,	,
zkie-zero-2020	274	14	meaning	mean	VBG
zkie-zero-2020	274	15	that	that	IN
zkie-zero-2020	274	16	we	-PRON-	PRP
zkie-zero-2020	274	17	have	have	VBP
zkie-zero-2020	274	18	ample	ample	JJ
zkie-zero-2020	274	19	evidence	evidence	NN
zkie-zero-2020	274	20	that	that	IN
zkie-zero-2020	274	21	our	-PRON-	PRP$
zkie-zero-2020	274	22	reranking	reranking	NN
zkie-zero-2020	274	23	speeds	speed	VBZ
zkie-zero-2020	274	24	up	up	RP
zkie-zero-2020	274	25	annotation	annotation	NN
zkie-zero-2020	274	26	time	time	NN
zkie-zero-2020	274	27	.	.	.
zkie-zero-2020	275	1	Recommender	recommender	NN
zkie-zero-2020	275	2	suggestions	suggestion	NNS
zkie-zero-2020	275	3	made	make	VBD
zkie-zero-2020	275	4	up	up	RP
zkie-zero-2020	275	5	around	around	IN
zkie-zero-2020	275	6	30	30	CD
zkie-zero-2020	275	7	%	%	NN
zkie-zero-2020	275	8	of	of	IN
zkie-zero-2020	275	9	annotations	annotation	NNS
zkie-zero-2020	275	10	.	.	.
zkie-zero-2020	276	1	We	-PRON-	PRP
zkie-zero-2020	276	2	did	do	VBD
zkie-zero-2020	276	3	not	not	RB
zkie-zero-2020	276	4	measure	measure	VB
zkie-zero-2020	276	5	a	a	DT
zkie-zero-2020	276	6	significant	significant	JJ
zkie-zero-2020	276	7	difference	difference	NN
zkie-zero-2020	276	8	between	between	IN
zkie-zero-2020	276	9	string	string	NN
zkie-zero-2020	276	10	and	and	CC
zkie-zero-2020	276	11	Levenshtein	Levenshtein	NNP
zkie-zero-2020	276	12	recom-	recom-	NN
zkie-zero-2020	276	13	mender	mender	NN
zkie-zero-2020	276	14	.	.	.
zkie-zero-2020	277	1	About	about	RB
zkie-zero-2020	277	2	the	the	DT
zkie-zero-2020	277	3	latter	latter	JJ
zkie-zero-2020	277	4	,	,	,
zkie-zero-2020	277	5	users	user	NNS
zkie-zero-2020	277	6	liked	like	VBD
zkie-zero-2020	277	7	that	that	IN
zkie-zero-2020	277	8	it	-PRON-	PRP
zkie-zero-2020	277	9	can	can	MD
zkie-zero-2020	277	10	suggest	suggest	VB
zkie-zero-2020	277	11	annotations	annotation	NNS
zkie-zero-2020	277	12	for	for	IN
zkie-zero-2020	277	13	inexact	inexact	NN
zkie-zero-2020	277	14	matches	match	NNS
zkie-zero-2020	277	15	.	.	.
zkie-zero-2020	278	1	How-	How-	NNP
zkie-zero-2020	278	2	ever	ever	RB
zkie-zero-2020	278	3	,	,	,
zkie-zero-2020	278	4	they	-PRON-	PRP
zkie-zero-2020	278	5	criticized	criticize	VBD
zkie-zero-2020	278	6	the	the	DT
zkie-zero-2020	278	7	noisier	noisy	JJR
zkie-zero-2020	278	8	suggestions	suggestion	NNS
zkie-zero-2020	278	9	,	,	,
zkie-zero-2020	278	10	espe-	espe-	NNP
zkie-zero-2020	278	11	cially	cially	RB
zkie-zero-2020	278	12	for	for	IN
zkie-zero-2020	278	13	shorter	short	JJR
zkie-zero-2020	278	14	mentions	mention	NNS
zkie-zero-2020	278	15	(	(	-LRB-
zkie-zero-2020	278	16	e.g.	e.g.	RB
zkie-zero-2020	279	1	annotating	annotate	VBG
zkie-zero-2020	279	2	joabe	joabe	NNP
zkie-zero-2020	279	3	(	(	-LRB-
zkie-zero-2020	279	4	a	a	DT
zkie-zero-2020	279	5	name	name	NN
zkie-zero-2020	279	6	)	)	-RRB-
zkie-zero-2020	279	7	yielded	yield	VBD
zkie-zero-2020	279	8	suggestions	suggestion	NNS
zkie-zero-2020	279	9	for	for	IN
zkie-zero-2020	279	10	to	to	TO
zkie-zero-2020	279	11	be	be	VB
zkie-zero-2020	279	12	)	)	-RRB-
zkie-zero-2020	279	13	.	.	.
zkie-zero-2020	280	1	In	in	IN
zkie-zero-2020	280	2	the	the	DT
zkie-zero-2020	280	3	future	future	NN
zkie-zero-2020	280	4	,	,	,
zkie-zero-2020	280	5	we	-PRON-	PRP
zkie-zero-2020	280	6	will	will	MD
zkie-zero-2020	280	7	address	address	VB
zkie-zero-2020	280	8	this	this	DT
zkie-zero-2020	280	9	issue	issue	NN
zkie-zero-2020	280	10	by	by	IN
zkie-zero-2020	280	11	filtering	filter	VBG
zkie-zero-2020	280	12	out	out	RP
zkie-zero-2020	280	13	more	more	JJR
zkie-zero-2020	280	14	potentially	potentially	RB
zkie-zero-2020	280	15	unhelpful	unhelpful	JJ
zkie-zero-2020	280	16	suggestions	suggestion	NNS
zkie-zero-2020	280	17	and	and	CC
zkie-zero-2020	280	18	using	use	VBG
zkie-zero-2020	280	19	annotation	annotation	NN
zkie-zero-2020	280	20	rejections	rejection	NNS
zkie-zero-2020	280	21	as	as	IN
zkie-zero-2020	280	22	a	a	DT
zkie-zero-2020	280	23	blacklist	blacklist	NN
zkie-zero-2020	280	24	.	.	.
zkie-zero-2020	281	1	6	6	CD
zkie-zero-2020	281	2	Conclusion	Conclusion	NNP
zkie-zero-2020	281	3	We	-PRON-	PRP
zkie-zero-2020	281	4	presented	present	VBD
zkie-zero-2020	281	5	a	a	DT
zkie-zero-2020	281	6	domain	domain	NN
zkie-zero-2020	281	7	-	-	HYPH
zkie-zero-2020	281	8	agnostic	agnostic	JJ
zkie-zero-2020	281	9	annotation	annotation	NN
zkie-zero-2020	281	10	ap-	ap-	IN
zkie-zero-2020	281	11	proach	proach	NN
zkie-zero-2020	281	12	for	for	IN
zkie-zero-2020	281	13	annotating	annotate	VBG
zkie-zero-2020	281	14	entity	entity	NN
zkie-zero-2020	281	15	linking	link	VBG
zkie-zero-2020	281	16	for	for	IN
zkie-zero-2020	281	17	low-	low-	NN
zkie-zero-2020	281	18	resource	resource	NN
zkie-zero-2020	281	19	domains	domain	NNS
zkie-zero-2020	281	20	.	.	.
zkie-zero-2020	282	1	It	-PRON-	PRP
zkie-zero-2020	282	2	consists	consist	VBZ
zkie-zero-2020	282	3	of	of	IN
zkie-zero-2020	282	4	two	two	CD
zkie-zero-2020	282	5	main	main	JJ
zkie-zero-2020	282	6	com-	com-	NN
zkie-zero-2020	282	7	https://inception-project.github.io	https://inception-project.github.io	UH
zkie-zero-2020	282	8	6990	6990	CD
zkie-zero-2020	282	9	Figure	figure	NN
zkie-zero-2020	282	10	5	5	CD
zkie-zero-2020	282	11	:	:	:
zkie-zero-2020	282	12	For	for	IN
zkie-zero-2020	282	13	our	-PRON-	PRP$
zkie-zero-2020	282	14	user	user	NN
zkie-zero-2020	282	15	study	study	NN
zkie-zero-2020	282	16	,	,	,
zkie-zero-2020	282	17	we	-PRON-	PRP
zkie-zero-2020	282	18	extend	extend	VBP
zkie-zero-2020	282	19	the	the	DT
zkie-zero-2020	282	20	INCEpTION	inception	CD
zkie-zero-2020	282	21	annotation	annotation	NN
zkie-zero-2020	282	22	framework	framework	NN
zkie-zero-2020	282	23	:	:	:
zkie-zero-2020	282	24	1	1	CD
zkie-zero-2020	282	25	©	©	CD
zkie-zero-2020	282	26	entity	entity	NN
zkie-zero-2020	282	27	linking	linking	NN
zkie-zero-2020	282	28	search	search	NN
zkie-zero-2020	282	29	field	field	NN
zkie-zero-2020	282	30	,	,	,
zkie-zero-2020	282	31	2	2	CD
zkie-zero-2020	282	32	©	©	NNP
zkie-zero-2020	282	33	candidate	candidate	NN
zkie-zero-2020	282	34	list	list	NN
zkie-zero-2020	282	35	,	,	,
zkie-zero-2020	282	36	3	3	CD
zkie-zero-2020	282	37	©	©	NNP
zkie-zero-2020	282	38	linked	link	VBN
zkie-zero-2020	282	39	named	name	VBN
zkie-zero-2020	282	40	entity	entity	NN
zkie-zero-2020	282	41	,	,	,
zkie-zero-2020	282	42	4	4	CD
zkie-zero-2020	282	43	©	©	NNP
zkie-zero-2020	282	44	entity	entity	NN
zkie-zero-2020	282	45	linking	link	VBG
zkie-zero-2020	282	46	recommendation	recommendation	NN
zkie-zero-2020	282	47	.	.	.
zkie-zero-2020	283	1	ponents	ponent	NNS
zkie-zero-2020	283	2	:	:	:
zkie-zero-2020	283	3	recommenders	recommender	NNS
zkie-zero-2020	283	4	that	that	WDT
zkie-zero-2020	283	5	are	be	VBP
zkie-zero-2020	283	6	algorithms	algorithm	NNS
zkie-zero-2020	283	7	that	that	WDT
zkie-zero-2020	283	8	suggest	suggest	VBP
zkie-zero-2020	283	9	potential	potential	JJ
zkie-zero-2020	283	10	annotations	annotation	NNS
zkie-zero-2020	283	11	to	to	IN
zkie-zero-2020	283	12	users	user	NNS
zkie-zero-2020	283	13	and	and	CC
zkie-zero-2020	283	14	a	a	DT
zkie-zero-2020	283	15	ranker	ranker	NN
zkie-zero-2020	283	16	that	that	IN
zkie-zero-2020	283	17	,	,	,
zkie-zero-2020	283	18	given	give	VBN
zkie-zero-2020	283	19	a	a	DT
zkie-zero-2020	283	20	mention	mention	NN
zkie-zero-2020	283	21	span	span	NN
zkie-zero-2020	283	22	,	,	,
zkie-zero-2020	283	23	ranks	rank	VBZ
zkie-zero-2020	283	24	potential	potential	JJ
zkie-zero-2020	283	25	entity	entity	NN
zkie-zero-2020	283	26	candidates	candidate	NNS
zkie-zero-2020	283	27	so	so	IN
zkie-zero-2020	283	28	that	that	IN
zkie-zero-2020	283	29	they	-PRON-	PRP
zkie-zero-2020	283	30	show	show	VBP
zkie-zero-2020	283	31	up	up	RP
zkie-zero-2020	283	32	higher	higher	RBR
zkie-zero-2020	283	33	in	in	IN
zkie-zero-2020	283	34	the	the	DT
zkie-zero-2020	283	35	can-	can-	NN
zkie-zero-2020	283	36	didate	didate	NN
zkie-zero-2020	283	37	list	list	NN
zkie-zero-2020	283	38	,	,	,
zkie-zero-2020	283	39	making	make	VBG
zkie-zero-2020	283	40	it	-PRON-	PRP
zkie-zero-2020	283	41	easier	easy	JJR
zkie-zero-2020	283	42	to	to	TO
zkie-zero-2020	283	43	find	find	VB
zkie-zero-2020	283	44	for	for	IN
zkie-zero-2020	283	45	users	user	NNS
zkie-zero-2020	283	46	.	.	.
zkie-zero-2020	284	1	Both	both	DT
zkie-zero-2020	284	2	systems	system	NNS
zkie-zero-2020	284	3	are	be	VBP
zkie-zero-2020	284	4	retrained	retrain	VBN
zkie-zero-2020	284	5	whenever	whenever	WRB
zkie-zero-2020	284	6	new	new	JJ
zkie-zero-2020	284	7	annotations	annotation	NNS
zkie-zero-2020	284	8	are	be	VBP
zkie-zero-2020	284	9	made	make	VBN
zkie-zero-2020	284	10	,	,	,
zkie-zero-2020	284	11	forming	form	VBG
zkie-zero-2020	284	12	the	the	DT
zkie-zero-2020	284	13	Human	Human	NNP
zkie-zero-2020	284	14	-	-	HYPH
zkie-zero-2020	284	15	In	in	IN
zkie-zero-2020	284	16	-	-	HYPH
zkie-zero-2020	284	17	The	The	NNP
zkie-zero-2020	284	18	-	-	HYPH
zkie-zero-2020	284	19	Loop	Loop	NNP
zkie-zero-2020	284	20	.	.	.
zkie-zero-2020	285	1	Our	-PRON-	PRP$
zkie-zero-2020	285	2	approach	approach	NN
zkie-zero-2020	285	3	does	do	VBZ
zkie-zero-2020	285	4	not	not	RB
zkie-zero-2020	285	5	require	require	VB
zkie-zero-2020	285	6	the	the	DT
zkie-zero-2020	285	7	existence	existence	NN
zkie-zero-2020	285	8	of	of	IN
zkie-zero-2020	285	9	external	external	JJ
zkie-zero-2020	285	10	resources	resource	NNS
zkie-zero-2020	285	11	like	like	IN
zkie-zero-2020	285	12	labeled	label	VBN
zkie-zero-2020	285	13	data	datum	NNS
zkie-zero-2020	285	14	,	,	,
zkie-zero-2020	285	15	tools	tool	NNS
zkie-zero-2020	285	16	like	like	IN
zkie-zero-2020	285	17	named	name	VBN
zkie-zero-2020	285	18	entity	entity	NN
zkie-zero-2020	285	19	recognizers	recognizer	NNS
zkie-zero-2020	285	20	or	or	CC
zkie-zero-2020	285	21	large	large	JJ
zkie-zero-2020	285	22	-	-	HYPH
zkie-zero-2020	285	23	scale	scale	NN
zkie-zero-2020	285	24	resources	resource	NNS
zkie-zero-2020	285	25	like	like	IN
zkie-zero-2020	285	26	Wikipedia	Wikipedia	NNP
zkie-zero-2020	285	27	.	.	.
zkie-zero-2020	286	1	It	-PRON-	PRP
zkie-zero-2020	286	2	can	can	MD
zkie-zero-2020	286	3	be	be	VB
zkie-zero-2020	286	4	applied	apply	VBN
zkie-zero-2020	286	5	to	to	IN
zkie-zero-2020	286	6	any	any	DT
zkie-zero-2020	286	7	domain	domain	NN
zkie-zero-2020	286	8	,	,	,
zkie-zero-2020	286	9	only	only	RB
zkie-zero-2020	286	10	requiring	require	VBG
zkie-zero-2020	286	11	a	a	DT
zkie-zero-2020	286	12	knowledge	knowledge	NN
zkie-zero-2020	286	13	base	base	NN
zkie-zero-2020	286	14	whose	whose	WP$
zkie-zero-2020	286	15	entities	entity	NNS
zkie-zero-2020	286	16	have	have	VBP
zkie-zero-2020	286	17	a	a	DT
zkie-zero-2020	286	18	label	label	NN
zkie-zero-2020	286	19	and	and	CC
zkie-zero-2020	286	20	a	a	DT
zkie-zero-2020	286	21	description	description	NN
zkie-zero-2020	286	22	.	.	.
zkie-zero-2020	287	1	In	in	IN
zkie-zero-2020	287	2	this	this	DT
zkie-zero-2020	287	3	paper	paper	NN
zkie-zero-2020	287	4	,	,	,
zkie-zero-2020	287	5	we	-PRON-	PRP
zkie-zero-2020	287	6	evaluate	evaluate	VBP
zkie-zero-2020	287	7	on	on	IN
zkie-zero-2020	287	8	three	three	CD
zkie-zero-2020	287	9	datasets	dataset	NNS
zkie-zero-2020	287	10	:	:	:
zkie-zero-2020	287	11	AIDA	AIDA	NNP
zkie-zero-2020	287	12	,	,	,
zkie-zero-2020	287	13	which	which	WDT
zkie-zero-2020	287	14	is	be	VBZ
zkie-zero-2020	287	15	often	often	RB
zkie-zero-2020	287	16	used	use	VBN
zkie-zero-2020	287	17	to	to	TO
zkie-zero-2020	287	18	validate	validate	VB
zkie-zero-2020	287	19	state	state	NN
zkie-zero-2020	287	20	-	-	HYPH
zkie-zero-2020	287	21	of	of	IN
zkie-zero-2020	287	22	-	-	HYPH
zkie-zero-2020	287	23	the	the	DT
zkie-zero-2020	287	24	-	-	HYPH
zkie-zero-2020	287	25	art	art	NN
zkie-zero-2020	287	26	entity	entity	NN
zkie-zero-2020	287	27	linking	link	VBG
zkie-zero-2020	287	28	sys-	sys-	NN
zkie-zero-2020	287	29	tems	tem	NNS
zkie-zero-2020	287	30	as	as	RB
zkie-zero-2020	287	31	well	well	RB
zkie-zero-2020	287	32	as	as	IN
zkie-zero-2020	287	33	WWO	wwo	NN
zkie-zero-2020	287	34	and	and	CC
zkie-zero-2020	287	35	1641	1641	CD
zkie-zero-2020	287	36	from	from	IN
zkie-zero-2020	287	37	the	the	DT
zkie-zero-2020	287	38	humanities	humanity	NNS
zkie-zero-2020	287	39	.	.	.
zkie-zero-2020	288	1	We	-PRON-	PRP
zkie-zero-2020	288	2	show	show	VBP
zkie-zero-2020	288	3	that	that	IN
zkie-zero-2020	288	4	in	in	IN
zkie-zero-2020	288	5	simulation	simulation	NN
zkie-zero-2020	288	6	,	,	,
zkie-zero-2020	288	7	only	only	RB
zkie-zero-2020	288	8	a	a	DT
zkie-zero-2020	288	9	very	very	RB
zkie-zero-2020	288	10	small	small	JJ
zkie-zero-2020	288	11	sub-	sub-	JJ
zkie-zero-2020	288	12	set	set	NN
zkie-zero-2020	288	13	needs	need	NNS
zkie-zero-2020	288	14	to	to	TO
zkie-zero-2020	288	15	be	be	VB
zkie-zero-2020	288	16	annotated	annotate	VBN
zkie-zero-2020	288	17	(	(	-LRB-
zkie-zero-2020	288	18	fewer	few	JJR
zkie-zero-2020	288	19	than	than	IN
zkie-zero-2020	288	20	100	100	CD
zkie-zero-2020	288	21	)	)	-RRB-
zkie-zero-2020	288	22	for	for	IN
zkie-zero-2020	288	23	the	the	DT
zkie-zero-2020	288	24	ranker	ranker	JJR
zkie-zero-2020	288	25	to	to	TO
zkie-zero-2020	288	26	reach	reach	VB
zkie-zero-2020	288	27	high	high	JJ
zkie-zero-2020	288	28	accuracy	accuracy	NN
zkie-zero-2020	288	29	.	.	.
zkie-zero-2020	289	1	In	in	IN
zkie-zero-2020	289	2	a	a	DT
zkie-zero-2020	289	3	user	user	NN
zkie-zero-2020	289	4	study	study	NN
zkie-zero-2020	289	5	,	,	,
zkie-zero-2020	289	6	re-	re-	JJ
zkie-zero-2020	289	7	sults	sult	NNS
zkie-zero-2020	289	8	show	show	VBP
zkie-zero-2020	289	9	that	that	IN
zkie-zero-2020	289	10	users	user	NNS
zkie-zero-2020	289	11	prefer	prefer	VBP
zkie-zero-2020	289	12	our	-PRON-	PRP$
zkie-zero-2020	289	13	approach	approach	NN
zkie-zero-2020	289	14	compared	compare	VBN
zkie-zero-2020	289	15	to	to	IN
zkie-zero-2020	289	16	the	the	DT
zkie-zero-2020	289	17	typical	typical	JJ
zkie-zero-2020	289	18	annotation	annotation	NN
zkie-zero-2020	289	19	process	process	NN
zkie-zero-2020	289	20	;	;	:
zkie-zero-2020	289	21	annotation	annotation	NN
zkie-zero-2020	289	22	speed	speed	NN
zkie-zero-2020	289	23	improves	improve	VBZ
zkie-zero-2020	289	24	by	by	IN
zkie-zero-2020	289	25	around	around	IN
zkie-zero-2020	289	26	35	35	CD
zkie-zero-2020	289	27	%	%	NN
zkie-zero-2020	289	28	when	when	WRB
zkie-zero-2020	289	29	using	use	VBG
zkie-zero-2020	289	30	our	-PRON-	PRP$
zkie-zero-2020	289	31	system	system	NN
zkie-zero-2020	289	32	relative	relative	JJ
zkie-zero-2020	289	33	to	to	IN
zkie-zero-2020	289	34	using	use	VBG
zkie-zero-2020	289	35	no	no	DT
zkie-zero-2020	289	36	reranking	reranking	NN
zkie-zero-2020	289	37	support	support	NN
zkie-zero-2020	289	38	.	.	.
zkie-zero-2020	290	1	In	in	IN
zkie-zero-2020	290	2	the	the	DT
zkie-zero-2020	290	3	future	future	NN
zkie-zero-2020	290	4	,	,	,
zkie-zero-2020	290	5	we	-PRON-	PRP
zkie-zero-2020	290	6	want	want	VBP
zkie-zero-2020	290	7	to	to	TO
zkie-zero-2020	290	8	investigate	investigate	VB
zkie-zero-2020	290	9	more	more	RBR
zkie-zero-2020	290	10	power-	power-	JJ
zkie-zero-2020	290	11	ful	ful	JJ
zkie-zero-2020	290	12	recommenders	recommender	NNS
zkie-zero-2020	290	13	,	,	,
zkie-zero-2020	290	14	combine	combine	VB
zkie-zero-2020	290	15	interactive	interactive	JJ
zkie-zero-2020	290	16	entity	entity	NN
zkie-zero-2020	290	17	link-	link-	NN
zkie-zero-2020	290	18	ing	e	VBG
zkie-zero-2020	290	19	with	with	IN
zkie-zero-2020	290	20	knowledge	knowledge	NN
zkie-zero-2020	290	21	base	base	NN
zkie-zero-2020	290	22	completion	completion	NN
zkie-zero-2020	290	23	and	and	CC
zkie-zero-2020	290	24	use	use	VB
zkie-zero-2020	290	25	online	online	JJ
zkie-zero-2020	290	26	learning	learning	NN
zkie-zero-2020	290	27	to	to	TO
zkie-zero-2020	290	28	leverage	leverage	NN
zkie-zero-2020	290	29	deep	deep	JJ
zkie-zero-2020	290	30	models	model	NNS
zkie-zero-2020	290	31	,	,	,
zkie-zero-2020	290	32	despite	despite	IN
zkie-zero-2020	290	33	their	-PRON-	PRP$
zkie-zero-2020	290	34	long	long	JJ
zkie-zero-2020	290	35	training	training	NN
zkie-zero-2020	290	36	time	time	NN
zkie-zero-2020	290	37	.	.	.
zkie-zero-2020	291	1	Acknowledgments	acknowledgment	NNS
zkie-zero-2020	291	2	We	-PRON-	PRP
zkie-zero-2020	291	3	thank	thank	VBP
zkie-zero-2020	291	4	the	the	DT
zkie-zero-2020	291	5	anonymous	anonymous	JJ
zkie-zero-2020	291	6	reviewers	reviewer	NNS
zkie-zero-2020	291	7	and	and	CC
zkie-zero-2020	291	8	Kevin	Kevin	NNP
zkie-zero-2020	291	9	Stowe	Stowe	NNP
zkie-zero-2020	291	10	for	for	IN
zkie-zero-2020	291	11	their	-PRON-	PRP$
zkie-zero-2020	291	12	detailed	detailed	JJ
zkie-zero-2020	291	13	and	and	CC
zkie-zero-2020	291	14	helpful	helpful	JJ
zkie-zero-2020	291	15	comments	comment	NNS
zkie-zero-2020	291	16	.	.	.
zkie-zero-2020	292	1	We	-PRON-	PRP
zkie-zero-2020	292	2	also	also	RB
zkie-zero-2020	292	3	want	want	VBP
zkie-zero-2020	292	4	to	to	TO
zkie-zero-2020	292	5	thank	thank	VB
zkie-zero-2020	292	6	the	the	DT
zkie-zero-2020	292	7	Women	Women	NNP
zkie-zero-2020	292	8	Writers	Writers	NNPS
zkie-zero-2020	292	9	Project	Project	NNP
zkie-zero-2020	292	10	which	which	WDT
zkie-zero-2020	292	11	made	make	VBD
zkie-zero-2020	292	12	the	the	DT
zkie-zero-2020	292	13	Women	Women	NNP
zkie-zero-2020	292	14	Writers	Writers	NNPS
zkie-zero-2020	292	15	Online	Online	NNP
zkie-zero-2020	292	16	text	text	NN
zkie-zero-2020	292	17	col-	col-	JJ
zkie-zero-2020	292	18	lection	lection	NN
zkie-zero-2020	292	19	available	available	JJ
zkie-zero-2020	292	20	to	to	IN
zkie-zero-2020	292	21	us	-PRON-	PRP
zkie-zero-2020	292	22	.	.	.
zkie-zero-2020	293	1	This	this	DT
zkie-zero-2020	293	2	work	work	NN
zkie-zero-2020	293	3	was	be	VBD
zkie-zero-2020	293	4	supported	support	VBN
zkie-zero-2020	293	5	by	by	IN
zkie-zero-2020	293	6	the	the	DT
zkie-zero-2020	293	7	German	German	NNP
zkie-zero-2020	293	8	Research	Research	NNP
zkie-zero-2020	293	9	Foundation	Foundation	NNP
zkie-zero-2020	293	10	under	under	IN
zkie-zero-2020	293	11	grant	grant	NN
zkie-zero-2020	293	12	№	№	NNP
zkie-zero-2020	293	13	EC	EC	NNP
zkie-zero-2020	293	14	503/1	503/1	NNP
zkie-zero-2020	293	15	-	-	HYPH
zkie-zero-2020	293	16	1	1	CD
zkie-zero-2020	293	17	and	and	CC
zkie-zero-2020	293	18	GU	GU	NNP
zkie-zero-2020	293	19	798/21	798/21	NNP
zkie-zero-2020	293	20	-	-	HYPH
zkie-zero-2020	293	21	1	1	CD
zkie-zero-2020	293	22	as	as	RB
zkie-zero-2020	293	23	well	well	RB
zkie-zero-2020	293	24	as	as	IN
zkie-zero-2020	293	25	by	by	IN
zkie-zero-2020	293	26	the	the	DT
zkie-zero-2020	293	27	German	German	NNP
zkie-zero-2020	293	28	Federal	Federal	NNP
zkie-zero-2020	293	29	Ministry	Ministry	NNP
zkie-zero-2020	293	30	of	of	IN
zkie-zero-2020	293	31	Education	Education	NNP
zkie-zero-2020	293	32	and	and	CC
zkie-zero-2020	293	33	Re-	Re-	NNP
zkie-zero-2020	293	34	search	search	NN
zkie-zero-2020	293	35	(	(	-LRB-
zkie-zero-2020	293	36	BMBF	BMBF	NNP
zkie-zero-2020	293	37	)	)	-RRB-
zkie-zero-2020	293	38	under	under	IN
zkie-zero-2020	293	39	the	the	DT
zkie-zero-2020	293	40	promotional	promotional	JJ
zkie-zero-2020	293	41	reference	reference	NN
zkie-zero-2020	293	42	01UG1816B	01UG1816B	.
zkie-zero-2020	293	43	(	(	-LRB-
zkie-zero-2020	293	44	CEDIFOR	CEDIFOR	NNP
zkie-zero-2020	293	45	)	)	-RRB-
zkie-zero-2020	293	46	.	.	.
zkie-zero-2020	294	1	References	reference	NNS
zkie-zero-2020	294	2	Sebastian	Sebastian	NNP
zkie-zero-2020	294	3	Arnold	Arnold	NNP
zkie-zero-2020	294	4	,	,	,
zkie-zero-2020	294	5	Robert	Robert	NNP
zkie-zero-2020	294	6	Dziuba	Dziuba	NNP
zkie-zero-2020	294	7	,	,	,
zkie-zero-2020	294	8	and	and	CC
zkie-zero-2020	294	9	Alexander	Alexander	NNP
zkie-zero-2020	294	10	Löser	Löser	NNP
zkie-zero-2020	294	11	.	.	.
zkie-zero-2020	295	1	2016	2016	CD
zkie-zero-2020	295	2	.	.	.
zkie-zero-2020	296	1	TASTY	TASTY	NNP
zkie-zero-2020	296	2	:	:	:
zkie-zero-2020	296	3	Interactive	interactive	JJ
zkie-zero-2020	296	4	Entity	entity	NN
zkie-zero-2020	296	5	Linking	Linking	NNP
zkie-zero-2020	296	6	As	As	NNP
zkie-zero-2020	296	7	-	-	HYPH
zkie-zero-2020	296	8	You-	You-	NNP
zkie-zero-2020	296	9	Type	type	NN
zkie-zero-2020	296	10	.	.	.
zkie-zero-2020	297	1	In	in	IN
zkie-zero-2020	297	2	Proceedings	Proceedings	NNP
zkie-zero-2020	297	3	of	of	IN
zkie-zero-2020	297	4	COLING	COLING	NNP
zkie-zero-2020	297	5	2016	2016	CD
zkie-zero-2020	297	6	,	,	,
zkie-zero-2020	297	7	the	the	DT
zkie-zero-2020	297	8	26th	26th	JJ
zkie-zero-2020	297	9	International	International	NNP
zkie-zero-2020	297	10	Conference	Conference	NNP
zkie-zero-2020	297	11	on	on	IN
zkie-zero-2020	297	12	Computational	Computational	NNP
zkie-zero-2020	297	13	Linguis-	Linguis-	NNP
zkie-zero-2020	297	14	tics	tic	NNS
zkie-zero-2020	297	15	:	:	:
zkie-zero-2020	297	16	System	System	NNP
zkie-zero-2020	297	17	Demonstrations	Demonstrations	NNPS
zkie-zero-2020	297	18	,	,	,
zkie-zero-2020	297	19	pages	page	VBZ
zkie-zero-2020	297	20	111–115	111–115	CD
zkie-zero-2020	297	21	.	.	.
zkie-zero-2020	298	1	Sabine	Sabine	NNP
zkie-zero-2020	298	2	Bartsch	Bartsch	NNP
zkie-zero-2020	298	3	.	.	.
zkie-zero-2020	299	1	2004	2004	CD
zkie-zero-2020	299	2	.	.	.
zkie-zero-2020	300	1	Annotating	annotate	VBG
zkie-zero-2020	300	2	a	a	DT
zkie-zero-2020	300	3	Corpus	Corpus	NNP
zkie-zero-2020	300	4	for	for	IN
zkie-zero-2020	300	5	Build-	Build-	NNP
zkie-zero-2020	300	6	ing	e	VBG
zkie-zero-2020	300	7	a	a	DT
zkie-zero-2020	300	8	Domain	domain	NN
zkie-zero-2020	300	9	-	-	HYPH
zkie-zero-2020	300	10	specific	specific	JJ
zkie-zero-2020	300	11	Knowledge	Knowledge	NNP
zkie-zero-2020	300	12	Base	Base	NNP
zkie-zero-2020	300	13	.	.	.
zkie-zero-2020	301	1	In	in	IN
zkie-zero-2020	301	2	Proceed-	Proceed-	NNP
zkie-zero-2020	301	3	ings	ing	NNS
zkie-zero-2020	301	4	of	of	IN
zkie-zero-2020	301	5	the	the	DT
zkie-zero-2020	301	6	Fourth	Fourth	NNP
zkie-zero-2020	301	7	International	International	NNP
zkie-zero-2020	301	8	Conference	Conference	NNP
zkie-zero-2020	301	9	on	on	IN
zkie-zero-2020	301	10	Lan-	Lan-	NNP
zkie-zero-2020	301	11	guage	guage	NN
zkie-zero-2020	301	12	Resources	Resources	NNPS
zkie-zero-2020	301	13	and	and	CC
zkie-zero-2020	301	14	Evaluation	Evaluation	NNP
zkie-zero-2020	301	15	(	(	-LRB-
zkie-zero-2020	301	16	LREC’04	LREC’04	NNP
zkie-zero-2020	301	17	)	)	-RRB-
zkie-zero-2020	301	18	,	,	,
zkie-zero-2020	301	19	pages	page	VBZ
zkie-zero-2020	301	20	1669–1672	1669–1672	CD
zkie-zero-2020	301	21	.	.	.
zkie-zero-2020	302	1	Carmen	Carmen	NNP
zkie-zero-2020	302	2	Brando	Brando	NNP
zkie-zero-2020	302	3	,	,	,
zkie-zero-2020	302	4	Francesca	Francesca	NNP
zkie-zero-2020	302	5	Frontini	Frontini	NNP
zkie-zero-2020	302	6	,	,	,
zkie-zero-2020	302	7	and	and	CC
zkie-zero-2020	302	8	Jean	Jean	NNP
zkie-zero-2020	302	9	-	-	HYPH
zkie-zero-2020	302	10	Gabriel	Gabriel	NNP
zkie-zero-2020	302	11	Ganascia	Ganascia	NNP
zkie-zero-2020	302	12	.	.	.
zkie-zero-2020	303	1	2016	2016	CD
zkie-zero-2020	303	2	.	.	.
zkie-zero-2020	304	1	REDEN	REDEN	NNP
zkie-zero-2020	304	2	:	:	:
zkie-zero-2020	304	3	Named	name	VBN
zkie-zero-2020	304	4	Entity	Entity	NNP
zkie-zero-2020	304	5	Linking	Linking	NNP
zkie-zero-2020	304	6	in	in	IN
zkie-zero-2020	304	7	Digital	Digital	NNP
zkie-zero-2020	304	8	Literary	Literary	NNP
zkie-zero-2020	304	9	Editions	Editions	NNPS
zkie-zero-2020	304	10	Using	use	VBG
zkie-zero-2020	304	11	Linked	Linked	NNP
zkie-zero-2020	304	12	Data	Data	NNP
zkie-zero-2020	304	13	Sets	set	NNS
zkie-zero-2020	304	14	.	.	.
zkie-zero-2020	305	1	Complex	Complex	NNP
zkie-zero-2020	305	2	Systems	Systems	NNPS
zkie-zero-2020	305	3	Informatics	Informatics	NNPS
zkie-zero-2020	305	4	and	and	CC
zkie-zero-2020	305	5	Modeling	Modeling	NNP
zkie-zero-2020	305	6	Quar-	Quar-	NNP
zkie-zero-2020	305	7	terly	terly	RB
zkie-zero-2020	305	8	,	,	,
zkie-zero-2020	305	9	(	(	-LRB-
zkie-zero-2020	305	10	7):60–80	7):60–80	CD
zkie-zero-2020	305	11	.	.	.
zkie-zero-2020	306	1	Chris	Chris	NNP
zkie-zero-2020	306	2	Burges	Burges	NNP
zkie-zero-2020	306	3	,	,	,
zkie-zero-2020	306	4	Tal	Tal	NNP
zkie-zero-2020	306	5	Shaked	Shaked	NNP
zkie-zero-2020	306	6	,	,	,
zkie-zero-2020	306	7	Erin	Erin	NNP
zkie-zero-2020	306	8	Renshaw	Renshaw	NNP
zkie-zero-2020	306	9	,	,	,
zkie-zero-2020	306	10	Ari	Ari	NNP
zkie-zero-2020	306	11	Lazier	Lazier	NNP
zkie-zero-2020	306	12	,	,	,
zkie-zero-2020	306	13	Matt	Matt	NNP
zkie-zero-2020	306	14	Deeds	Deeds	NNP
zkie-zero-2020	306	15	,	,	,
zkie-zero-2020	306	16	Nicole	Nicole	NNP
zkie-zero-2020	306	17	Hamilton	Hamilton	NNP
zkie-zero-2020	306	18	,	,	,
zkie-zero-2020	306	19	and	and	CC
zkie-zero-2020	306	20	Greg	Greg	NNP
zkie-zero-2020	306	21	Hullender	Hullender	NNP
zkie-zero-2020	306	22	.	.	.
zkie-zero-2020	307	1	2005	2005	CD
zkie-zero-2020	307	2	.	.	.
zkie-zero-2020	308	1	Learning	learn	VBG
zkie-zero-2020	308	2	to	to	IN
zkie-zero-2020	308	3	rank	rank	NN
zkie-zero-2020	308	4	using	use	VBG
zkie-zero-2020	308	5	Gradient	Gradient	NNP
zkie-zero-2020	308	6	Descent	Descent	NNP
zkie-zero-2020	308	7	.	.	.
zkie-zero-2020	309	1	In	in	IN
zkie-zero-2020	309	2	Proceedings	proceeding	NNS
zkie-zero-2020	309	3	of	of	IN
zkie-zero-2020	309	4	the	the	DT
zkie-zero-2020	309	5	22nd	22nd	JJ
zkie-zero-2020	309	6	international	international	JJ
zkie-zero-2020	309	7	conference	conference	NN
zkie-zero-2020	309	8	on	on	IN
zkie-zero-2020	309	9	Machine	Machine	NNP
zkie-zero-2020	309	10	learning	learning	NN
zkie-zero-2020	309	11	-	-	HYPH
zkie-zero-2020	309	12	ICML	ICML	NNP
zkie-zero-2020	309	13	’	'	''
zkie-zero-2020	309	14	05	05	CD
zkie-zero-2020	309	15	,	,	,
zkie-zero-2020	309	16	pages	page	NNS
zkie-zero-2020	309	17	89–96	89–96	CD
zkie-zero-2020	309	18	.	.	.
zkie-zero-2020	310	1	Diego	Diego	NNP
zkie-zero-2020	310	2	Ceccarelli	Ceccarelli	NNP
zkie-zero-2020	310	3	,	,	,
zkie-zero-2020	310	4	Claudio	Claudio	NNP
zkie-zero-2020	310	5	Lucchese	Lucchese	NNP
zkie-zero-2020	310	6	,	,	,
zkie-zero-2020	310	7	Salvatore	Salvatore	NNP
zkie-zero-2020	310	8	Or-	or-	CC
zkie-zero-2020	310	9	lando	lando	VBZ
zkie-zero-2020	310	10	,	,	,
zkie-zero-2020	310	11	Raffaele	Raffaele	NNP
zkie-zero-2020	310	12	Perego	Perego	NNP
zkie-zero-2020	310	13	,	,	,
zkie-zero-2020	310	14	and	and	CC
zkie-zero-2020	310	15	Salvatore	Salvatore	NNP
zkie-zero-2020	310	16	Trani	Trani	NNP
zkie-zero-2020	310	17	.	.	.
zkie-zero-2020	311	1	2013	2013	CD
zkie-zero-2020	311	2	.	.	.
zkie-zero-2020	312	1	Dexter	dexter	NN
zkie-zero-2020	312	2	.	.	.
zkie-zero-2020	313	1	In	in	IN
zkie-zero-2020	313	2	Proceedings	Proceedings	NNP
zkie-zero-2020	313	3	of	of	IN
zkie-zero-2020	313	4	the	the	DT
zkie-zero-2020	313	5	sixth	sixth	JJ
zkie-zero-2020	313	6	international	international	JJ
zkie-zero-2020	313	7	workshop	workshop	NN
zkie-zero-2020	313	8	on	on	IN
zkie-zero-2020	313	9	Exploiting	exploit	VBG
zkie-zero-2020	313	10	semantic	semantic	JJ
zkie-zero-2020	313	11	annotations	annotation	NNS
zkie-zero-2020	313	12	in	in	IN
zkie-zero-2020	313	13	in-	in-	JJ
zkie-zero-2020	313	14	formation	formation	NN
zkie-zero-2020	313	15	retrieval	retrieval	NN
zkie-zero-2020	313	16	-	-	HYPH
zkie-zero-2020	313	17	ESAIR	ESAIR	NNP
zkie-zero-2020	313	18	'	'	POS
zkie-zero-2020	313	19	13	13	CD
zkie-zero-2020	313	20	,	,	,
zkie-zero-2020	313	21	pages	page	NNS
zkie-zero-2020	313	22	17–20	17–20	JJ
zkie-zero-2020	313	23	.	.	.
zkie-zero-2020	314	1	https://www.aclweb.org/anthology/C16-2024	https://www.aclweb.org/anthology/c16-2024	ADD
zkie-zero-2020	314	2	https://www.aclweb.org/anthology/C16-2024	https://www.aclweb.org/anthology/C16-2024	NNP
zkie-zero-2020	314	3	http://www.lrec-conf.org/proceedings/lrec2004/pdf/361.pdf	http://www.lrec-conf.org/proceedings/lrec2004/pdf/361.pdf	NNP
zkie-zero-2020	314	4	http://www.lrec-conf.org/proceedings/lrec2004/pdf/361.pdf	http://www.lrec-conf.org/proceedings/lrec2004/pdf/361.pdf	NNP
zkie-zero-2020	314	5	https://doi.org/10.7250/csimq.2016-7.04	https://doi.org/10.7250/csimq.2016-7.04	NNP
zkie-zero-2020	314	6	https://doi.org/10.7250/csimq.2016-7.04	https://doi.org/10.7250/csimq.2016-7.04	NNP
zkie-zero-2020	314	7	https://doi.org/10.1145/1102351.1102363	https://doi.org/10.1145/1102351.1102363	ADD
zkie-zero-2020	314	8	https://doi.org/10.1145/2513204.2513212	https://doi.org/10.1145/2513204.2513212	ADD
zkie-zero-2020	314	9	6991	6991	CD
zkie-zero-2020	314	10	Koby	Koby	NNP
zkie-zero-2020	314	11	Crammer	Crammer	NNP
zkie-zero-2020	314	12	and	and	CC
zkie-zero-2020	314	13	Yoram	Yoram	NNP
zkie-zero-2020	314	14	Singer	Singer	NNP
zkie-zero-2020	314	15	.	.	.
zkie-zero-2020	315	1	2003	2003	CD
zkie-zero-2020	315	2	.	.	.
zkie-zero-2020	316	1	Ultraconser-	Ultraconser-	NNP
zkie-zero-2020	316	2	vative	vative	JJ
zkie-zero-2020	316	3	Online	Online	NNP
zkie-zero-2020	316	4	Algorithms	Algorithms	NNP
zkie-zero-2020	316	5	for	for	IN
zkie-zero-2020	316	6	Multiclass	Multiclass	NNP
zkie-zero-2020	316	7	Problems	Problems	NNPS
zkie-zero-2020	316	8	.	.	.
zkie-zero-2020	317	1	JMLR	JMLR	NNP
zkie-zero-2020	317	2	,	,	,
zkie-zero-2020	317	3	3:951–991	3:951–991	CD
zkie-zero-2020	317	4	.	.	.
zkie-zero-2020	318	1	Joachim	Joachim	NNP
zkie-zero-2020	318	2	Daiber	Daiber	NNP
zkie-zero-2020	318	3	,	,	,
zkie-zero-2020	318	4	Max	Max	NNP
zkie-zero-2020	318	5	Jakob	Jakob	NNP
zkie-zero-2020	318	6	,	,	,
zkie-zero-2020	318	7	Chris	Chris	NNP
zkie-zero-2020	318	8	Hokamp	Hokamp	NNP
zkie-zero-2020	318	9	,	,	,
zkie-zero-2020	318	10	and	and	CC
zkie-zero-2020	318	11	Pablo	Pablo	NNP
zkie-zero-2020	318	12	N.	N.	NNP
zkie-zero-2020	318	13	Mendes	Mendes	NNP
zkie-zero-2020	318	14	.	.	.
zkie-zero-2020	319	1	2013	2013	CD
zkie-zero-2020	319	2	.	.	.
zkie-zero-2020	320	1	Improving	improve	VBG
zkie-zero-2020	320	2	efficiency	efficiency	NN
zkie-zero-2020	320	3	and	and	CC
zkie-zero-2020	320	4	accuracy	accuracy	NN
zkie-zero-2020	320	5	in	in	IN
zkie-zero-2020	320	6	multilingual	multilingual	JJ
zkie-zero-2020	320	7	entity	entity	NN
zkie-zero-2020	320	8	extraction	extraction	NN
zkie-zero-2020	320	9	.	.	.
zkie-zero-2020	321	1	In	in	IN
zkie-zero-2020	321	2	Pro-	Pro-	NNP
zkie-zero-2020	321	3	ceedings	ceeding	NNS
zkie-zero-2020	321	4	of	of	IN
zkie-zero-2020	321	5	the	the	DT
zkie-zero-2020	321	6	9th	9th	JJ
zkie-zero-2020	321	7	International	International	NNP
zkie-zero-2020	321	8	Conference	Conference	NNP
zkie-zero-2020	321	9	on	on	IN
zkie-zero-2020	321	10	Se-	Se-	NNP
zkie-zero-2020	321	11	mantic	mantic	JJ
zkie-zero-2020	321	12	Systems	Systems	NNPS
zkie-zero-2020	321	13	-	-	HYPH
zkie-zero-2020	321	14	I	I	NNP
zkie-zero-2020	321	15	-	-	HYPH
zkie-zero-2020	321	16	SEMANTICS	SEMANTICS	NNP
zkie-zero-2020	321	17	'	'	POS
zkie-zero-2020	321	18	13	13	CD
zkie-zero-2020	321	19	,	,	,
zkie-zero-2020	321	20	pages	page	VBZ
zkie-zero-2020	321	21	121–124	121–124	CD
zkie-zero-2020	321	22	.	.	.
zkie-zero-2020	322	1	Yadolah	Yadolah	NNP
zkie-zero-2020	322	2	Dodge	Dodge	NNP
zkie-zero-2020	322	3	.	.	.
zkie-zero-2020	323	1	2008	2008	CD
zkie-zero-2020	323	2	.	.	.
zkie-zero-2020	324	1	The	the	DT
zkie-zero-2020	324	2	Concise	Concise	NNP
zkie-zero-2020	324	3	Encyclopedia	Encyclopedia	NNP
zkie-zero-2020	324	4	of	of	IN
zkie-zero-2020	324	5	Statistics	Statistics	NNPS
zkie-zero-2020	324	6	.	.	.
zkie-zero-2020	325	1	Springer	Springer	NNP
zkie-zero-2020	325	2	.	.	.
zkie-zero-2020	326	1	Alexander	Alexander	NNP
zkie-zero-2020	326	2	Erdmann	Erdmann	NNP
zkie-zero-2020	326	3	,	,	,
zkie-zero-2020	326	4	David	David	NNP
zkie-zero-2020	326	5	Joseph	Joseph	NNP
zkie-zero-2020	326	6	Wrisley	Wrisley	NNP
zkie-zero-2020	326	7	,	,	,
zkie-zero-2020	326	8	Benjamin	Benjamin	NNP
zkie-zero-2020	326	9	Allen	Allen	NNP
zkie-zero-2020	326	10	,	,	,
zkie-zero-2020	326	11	Christopher	Christopher	NNP
zkie-zero-2020	326	12	Brown	Brown	NNP
zkie-zero-2020	326	13	,	,	,
zkie-zero-2020	326	14	Sophie	Sophie	NNP
zkie-zero-2020	326	15	Cohen	Cohen	NNP
zkie-zero-2020	326	16	-	-	HYPH
zkie-zero-2020	326	17	Bodénès	Bodénès	NNP
zkie-zero-2020	326	18	,	,	,
zkie-zero-2020	326	19	Micha	Micha	NNP
zkie-zero-2020	326	20	Elsner	Elsner	NNP
zkie-zero-2020	326	21	,	,	,
zkie-zero-2020	326	22	Yukun	Yukun	NNP
zkie-zero-2020	326	23	Feng	Feng	NNP
zkie-zero-2020	326	24	,	,	,
zkie-zero-2020	326	25	Brian	Brian	NNP
zkie-zero-2020	326	26	Joseph	Joseph	NNP
zkie-zero-2020	326	27	,	,	,
zkie-zero-2020	326	28	Béatrice	Béatrice	NNP
zkie-zero-2020	326	29	Joyeux	Joyeux	NNP
zkie-zero-2020	326	30	-	-	HYPH
zkie-zero-2020	326	31	Prunel	Prunel	NNP
zkie-zero-2020	326	32	,	,	,
zkie-zero-2020	326	33	and	and	CC
zkie-zero-2020	326	34	Marie	Marie	NNP
zkie-zero-2020	326	35	-	-	HYPH
zkie-zero-2020	326	36	Catherine	Catherine	NNP
zkie-zero-2020	326	37	de	de	FW
zkie-zero-2020	326	38	Marneffe	Marneffe	NNP
zkie-zero-2020	326	39	.	.	.
zkie-zero-2020	327	1	2019	2019	CD
zkie-zero-2020	327	2	.	.	.
zkie-zero-2020	328	1	Practical	practical	JJ
zkie-zero-2020	328	2	,	,	,
zkie-zero-2020	328	3	Efficient	Efficient	NNP
zkie-zero-2020	328	4	,	,	,
zkie-zero-2020	328	5	and	and	CC
zkie-zero-2020	328	6	Customizable	customizable	JJ
zkie-zero-2020	328	7	Active	Active	NNP
zkie-zero-2020	328	8	Learning	Learning	NNP
zkie-zero-2020	328	9	for	for	IN
zkie-zero-2020	328	10	Named	name	VBN
zkie-zero-2020	328	11	Entity	Entity	NNP
zkie-zero-2020	328	12	Recognition	Recognition	NNP
zkie-zero-2020	328	13	in	in	IN
zkie-zero-2020	328	14	the	the	DT
zkie-zero-2020	328	15	Digi-	Digi-	NNP
zkie-zero-2020	328	16	tal	tal	NN
zkie-zero-2020	328	17	Humanities	Humanities	NNPS
zkie-zero-2020	328	18	.	.	.
zkie-zero-2020	329	1	In	in	IN
zkie-zero-2020	329	2	Proceedings	Proceedings	NNP
zkie-zero-2020	329	3	of	of	IN
zkie-zero-2020	329	4	the	the	DT
zkie-zero-2020	329	5	2019	2019	CD
zkie-zero-2020	329	6	Confer-	Confer-	NNP
zkie-zero-2020	329	7	ence	ence	NN
zkie-zero-2020	329	8	of	of	IN
zkie-zero-2020	329	9	the	the	DT
zkie-zero-2020	329	10	North	North	NNP
zkie-zero-2020	329	11	,	,	,
zkie-zero-2020	329	12	pages	page	VBZ
zkie-zero-2020	329	13	2223–2234	2223–2234	CD
zkie-zero-2020	329	14	.	.	.
zkie-zero-2020	330	1	Paolo	Paolo	NNP
zkie-zero-2020	330	2	Ferragina	Ferragina	NNP
zkie-zero-2020	330	3	and	and	CC
zkie-zero-2020	330	4	Ugo	Ugo	NNP
zkie-zero-2020	330	5	Scaiella	Scaiella	NNP
zkie-zero-2020	330	6	.	.	.
zkie-zero-2020	331	1	2010	2010	CD
zkie-zero-2020	331	2	.	.	.
zkie-zero-2020	332	1	TAGME	tagme	RB
zkie-zero-2020	332	2	:	:	:
zkie-zero-2020	332	3	On	on	IN
zkie-zero-2020	332	4	-	-	HYPH
zkie-zero-2020	332	5	the	the	DT
zkie-zero-2020	332	6	-	-	HYPH
zkie-zero-2020	332	7	fly	fly	NN
zkie-zero-2020	332	8	Annotation	annotation	NN
zkie-zero-2020	332	9	of	of	IN
zkie-zero-2020	332	10	Short	Short	NNP
zkie-zero-2020	332	11	Text	Text	NNP
zkie-zero-2020	332	12	Fragments	Fragments	NNPS
zkie-zero-2020	332	13	(	(	-LRB-
zkie-zero-2020	332	14	by	by	IN
zkie-zero-2020	332	15	Wikipedia	Wikipedia	NNP
zkie-zero-2020	332	16	Entities	Entities	NNPS
zkie-zero-2020	332	17	)	)	-RRB-
zkie-zero-2020	332	18	.	.	.
zkie-zero-2020	333	1	In	in	IN
zkie-zero-2020	333	2	Proceedings	proceeding	NNS
zkie-zero-2020	333	3	of	of	IN
zkie-zero-2020	333	4	the	the	DT
zkie-zero-2020	333	5	19th	19th	JJ
zkie-zero-2020	333	6	ACM	ACM	NNP
zkie-zero-2020	333	7	international	international	JJ
zkie-zero-2020	333	8	conference	conference	NN
zkie-zero-2020	333	9	on	on	IN
zkie-zero-2020	333	10	Information	Information	NNP
zkie-zero-2020	333	11	and	and	CC
zkie-zero-2020	333	12	knowledge	knowledge	NN
zkie-zero-2020	333	13	management	management	NN
zkie-zero-2020	333	14	-	-	HYPH
zkie-zero-2020	333	15	CIKM	CIKM	NNP
zkie-zero-2020	333	16	'	'	POS
zkie-zero-2020	333	17	10	10	CD
zkie-zero-2020	333	18	,	,	,
zkie-zero-2020	333	19	pages	page	NNS
zkie-zero-2020	333	20	1625	1625	CD
zkie-zero-2020	333	21	–	–	:
zkie-zero-2020	333	22	1628	1628	CD
zkie-zero-2020	333	23	.	.	.
zkie-zero-2020	334	1	Yang	Yang	NNP
zkie-zero-2020	334	2	Gao	Gao	NNP
zkie-zero-2020	334	3	,	,	,
zkie-zero-2020	334	4	Christian	Christian	NNP
zkie-zero-2020	334	5	M.	M.	NNP
zkie-zero-2020	334	6	Meyer	Meyer	NNP
zkie-zero-2020	334	7	,	,	,
zkie-zero-2020	334	8	and	and	CC
zkie-zero-2020	334	9	Iryna	Iryna	NNP
zkie-zero-2020	334	10	Gurevych	Gurevych	NNP
zkie-zero-2020	334	11	.	.	.
zkie-zero-2020	335	1	2018	2018	CD
zkie-zero-2020	335	2	.	.	.
zkie-zero-2020	336	1	APRIL	APRIL	NNP
zkie-zero-2020	336	2	:	:	:
zkie-zero-2020	336	3	Interactively	interactively	RB
zkie-zero-2020	336	4	Learning	learn	VBG
zkie-zero-2020	336	5	to	to	IN
zkie-zero-2020	336	6	Summarise	Summarise	NNP
zkie-zero-2020	336	7	by	by	IN
zkie-zero-2020	336	8	Combining	combine	VBG
zkie-zero-2020	336	9	Active	active	JJ
zkie-zero-2020	336	10	Preference	Preference	NNP
zkie-zero-2020	336	11	Learning	Learning	NNP
zkie-zero-2020	336	12	and	and	CC
zkie-zero-2020	336	13	Re-	Re-	NNP
zkie-zero-2020	336	14	inforcement	inforcement	NN
zkie-zero-2020	336	15	Learning	learning	NN
zkie-zero-2020	336	16	.	.	.
zkie-zero-2020	337	1	In	in	IN
zkie-zero-2020	337	2	Proceedings	proceeding	NNS
zkie-zero-2020	337	3	of	of	IN
zkie-zero-2020	337	4	the	the	DT
zkie-zero-2020	337	5	2018	2018	CD
zkie-zero-2020	337	6	Conference	Conference	NNP
zkie-zero-2020	337	7	on	on	IN
zkie-zero-2020	337	8	Empirical	Empirical	NNP
zkie-zero-2020	337	9	Methods	Methods	NNPS
zkie-zero-2020	337	10	in	in	IN
zkie-zero-2020	337	11	Natural	Natural	NNP
zkie-zero-2020	337	12	Lan-	Lan-	NNP
zkie-zero-2020	337	13	guage	guage	NN
zkie-zero-2020	337	14	Processing	processing	NN
zkie-zero-2020	337	15	,	,	,
zkie-zero-2020	337	16	pages	page	VBZ
zkie-zero-2020	337	17	4120–4130	4120–4130	CD
zkie-zero-2020	337	18	.	.	.
zkie-zero-2020	338	1	Stephen	Stephen	NNP
zkie-zero-2020	338	2	Guo	Guo	NNP
zkie-zero-2020	338	3	,	,	,
zkie-zero-2020	338	4	Ming	Ming	NNP
zkie-zero-2020	338	5	-	-	HYPH
zkie-zero-2020	338	6	Wei	Wei	NNP
zkie-zero-2020	338	7	Chang	Chang	NNP
zkie-zero-2020	338	8	,	,	,
zkie-zero-2020	338	9	and	and	CC
zkie-zero-2020	338	10	Emre	Emre	NNP
zkie-zero-2020	338	11	Kiciman	Kiciman	NNP
zkie-zero-2020	338	12	.	.	.
zkie-zero-2020	339	1	2013	2013	CD
zkie-zero-2020	339	2	.	.	.
zkie-zero-2020	340	1	To	to	IN
zkie-zero-2020	340	2	Link	Link	NNP
zkie-zero-2020	340	3	or	or	CC
zkie-zero-2020	340	4	Not	not	RB
zkie-zero-2020	340	5	to	to	TO
zkie-zero-2020	340	6	Link	link	VB
zkie-zero-2020	340	7	?	?	.
zkie-zero-2020	341	1	A	a	DT
zkie-zero-2020	341	2	Study	Study	NNP
zkie-zero-2020	341	3	on	on	IN
zkie-zero-2020	341	4	End-	End-	NNP
zkie-zero-2020	341	5	to	to	TO
zkie-zero-2020	341	6	-	-	HYPH
zkie-zero-2020	341	7	End	end	NN
zkie-zero-2020	341	8	Tweet	Tweet	NNP
zkie-zero-2020	341	9	Entity	Entity	NNP
zkie-zero-2020	341	10	Linking	linking	NN
zkie-zero-2020	341	11	.	.	.
zkie-zero-2020	342	1	In	in	IN
zkie-zero-2020	342	2	Proceedings	Proceedings	NNP
zkie-zero-2020	342	3	of	of	IN
zkie-zero-2020	342	4	the	the	DT
zkie-zero-2020	342	5	2013	2013	CD
zkie-zero-2020	342	6	Conference	Conference	NNP
zkie-zero-2020	342	7	of	of	IN
zkie-zero-2020	342	8	the	the	DT
zkie-zero-2020	342	9	North	north	JJ
zkie-zero-2020	342	10	American	American	NNP
zkie-zero-2020	342	11	Chapter	Chapter	NNP
zkie-zero-2020	342	12	of	of	IN
zkie-zero-2020	342	13	the	the	DT
zkie-zero-2020	342	14	Association	Association	NNP
zkie-zero-2020	342	15	for	for	IN
zkie-zero-2020	342	16	Computational	Computational	NNP
zkie-zero-2020	342	17	Linguistics	Linguistics	NNPS
zkie-zero-2020	342	18	:	:	:
zkie-zero-2020	342	19	Hu-	Hu-	NNP
zkie-zero-2020	342	20	man	man	NN
zkie-zero-2020	342	21	Language	Language	NNP
zkie-zero-2020	342	22	Technologies	Technologies	NNPS
zkie-zero-2020	342	23	,	,	,
zkie-zero-2020	342	24	pages	page	VBZ
zkie-zero-2020	342	25	1020–1030	1020–1030	CD
zkie-zero-2020	342	26	.	.	.
zkie-zero-2020	343	1	Isabelle	Isabelle	NNP
zkie-zero-2020	343	2	Guyon	Guyon	NNP
zkie-zero-2020	343	3	,	,	,
zkie-zero-2020	343	4	Jason	Jason	NNP
zkie-zero-2020	343	5	Weston	Weston	NNP
zkie-zero-2020	343	6	,	,	,
zkie-zero-2020	343	7	Stephen	Stephen	NNP
zkie-zero-2020	343	8	Barnhill	Barnhill	NNP
zkie-zero-2020	343	9	,	,	,
zkie-zero-2020	343	10	and	and	CC
zkie-zero-2020	343	11	Vladimir	Vladimir	NNP
zkie-zero-2020	343	12	Vapnik	Vapnik	NNP
zkie-zero-2020	343	13	.	.	.
zkie-zero-2020	344	1	2002	2002	CD
zkie-zero-2020	344	2	.	.	.
zkie-zero-2020	345	1	Gene	Gene	NNP
zkie-zero-2020	345	2	Selection	Selection	NNP
zkie-zero-2020	345	3	for	for	IN
zkie-zero-2020	345	4	Cancer	Cancer	NNP
zkie-zero-2020	345	5	Classification	Classification	NNP
zkie-zero-2020	345	6	using	use	VBG
zkie-zero-2020	345	7	Support	Support	NNP
zkie-zero-2020	345	8	Vector	Vector	NNP
zkie-zero-2020	345	9	Machines	Machines	NNPS
zkie-zero-2020	345	10	.	.	.
zkie-zero-2020	346	1	Ma-	ma-	DT
zkie-zero-2020	346	2	chine	chine	NN
zkie-zero-2020	346	3	Learning	Learning	NNP
zkie-zero-2020	346	4	,	,	,
zkie-zero-2020	346	5	46:389–422	46:389–422	CD
zkie-zero-2020	346	6	.	.	.
zkie-zero-2020	347	1	Luheng	Luheng	NNP
zkie-zero-2020	347	2	He	He	NNP
zkie-zero-2020	347	3	,	,	,
zkie-zero-2020	347	4	Julian	Julian	NNP
zkie-zero-2020	347	5	Michael	Michael	NNP
zkie-zero-2020	347	6	,	,	,
zkie-zero-2020	347	7	Mike	Mike	NNP
zkie-zero-2020	347	8	Lewis	Lewis	NNP
zkie-zero-2020	347	9	,	,	,
zkie-zero-2020	347	10	and	and	CC
zkie-zero-2020	347	11	Luke	Luke	NNP
zkie-zero-2020	347	12	Zettlemoyer	Zettlemoyer	NNP
zkie-zero-2020	347	13	.	.	.
zkie-zero-2020	348	1	2016	2016	CD
zkie-zero-2020	348	2	.	.	.
zkie-zero-2020	349	1	Human	human	JJ
zkie-zero-2020	349	2	-	-	HYPH
zkie-zero-2020	349	3	in	in	IN
zkie-zero-2020	349	4	-	-	HYPH
zkie-zero-2020	349	5	the	the	DT
zkie-zero-2020	349	6	-	-	HYPH
zkie-zero-2020	349	7	Loop	Loop	NNP
zkie-zero-2020	349	8	Parsing	Parsing	NNP
zkie-zero-2020	349	9	.	.	.
zkie-zero-2020	350	1	In	in	IN
zkie-zero-2020	350	2	Proceedings	proceeding	NNS
zkie-zero-2020	350	3	of	of	IN
zkie-zero-2020	350	4	the	the	DT
zkie-zero-2020	350	5	2016	2016	CD
zkie-zero-2020	350	6	Conference	Conference	NNP
zkie-zero-2020	350	7	on	on	IN
zkie-zero-2020	350	8	Empirical	Empirical	NNP
zkie-zero-2020	350	9	Methods	Methods	NNPS
zkie-zero-2020	350	10	in	in	IN
zkie-zero-2020	350	11	Natural	Natural	NNP
zkie-zero-2020	350	12	Language	Language	NNP
zkie-zero-2020	350	13	Processing	Processing	NNP
zkie-zero-2020	350	14	,	,	,
zkie-zero-2020	350	15	pages	page	NNS
zkie-zero-2020	350	16	2337–2342	2337–2342	CD
zkie-zero-2020	350	17	.	.	.
zkie-zero-2020	351	1	Johannes	Johannes	NNP
zkie-zero-2020	351	2	Hoffart	Hoffart	NNP
zkie-zero-2020	351	3	,	,	,
zkie-zero-2020	351	4	Mohamed	Mohamed	NNP
zkie-zero-2020	351	5	Amir	Amir	NNP
zkie-zero-2020	351	6	Yosef	Yosef	NNP
zkie-zero-2020	351	7	,	,	,
zkie-zero-2020	351	8	Ilaria	Ilaria	NNP
zkie-zero-2020	351	9	Bor-	Bor-	NNP
zkie-zero-2020	351	10	dino	dino	NN
zkie-zero-2020	351	11	,	,	,
zkie-zero-2020	351	12	Hagen	Hagen	NNP
zkie-zero-2020	351	13	Fürstenau	Fürstenau	NNP
zkie-zero-2020	351	14	,	,	,
zkie-zero-2020	351	15	Manfred	Manfred	NNP
zkie-zero-2020	351	16	Pinkal	Pinkal	NNP
zkie-zero-2020	351	17	,	,	,
zkie-zero-2020	351	18	Marc	Marc	NNP
zkie-zero-2020	351	19	Span-	Span-	NNP
zkie-zero-2020	351	20	iol	iol	NNP
zkie-zero-2020	351	21	,	,	,
zkie-zero-2020	351	22	Bilyana	Bilyana	NNP
zkie-zero-2020	351	23	Taneva	Taneva	NNP
zkie-zero-2020	351	24	,	,	,
zkie-zero-2020	351	25	Stefan	Stefan	NNP
zkie-zero-2020	351	26	Thater	Thater	NNP
zkie-zero-2020	351	27	,	,	,
zkie-zero-2020	351	28	and	and	CC
zkie-zero-2020	351	29	Gerhard	Gerhard	NNP
zkie-zero-2020	351	30	Weikum	Weikum	NNP
zkie-zero-2020	351	31	.	.	.
zkie-zero-2020	352	1	2011	2011	CD
zkie-zero-2020	352	2	.	.	.
zkie-zero-2020	353	1	Robust	robust	JJ
zkie-zero-2020	353	2	Disambiguation	Disambiguation	NNP
zkie-zero-2020	353	3	of	of	IN
zkie-zero-2020	353	4	Named	Named	NNP
zkie-zero-2020	353	5	Entities	Entities	NNP
zkie-zero-2020	353	6	in	in	IN
zkie-zero-2020	353	7	Text	Text	NNP
zkie-zero-2020	353	8	.	.	.
zkie-zero-2020	354	1	In	in	IN
zkie-zero-2020	354	2	Proceedings	Proceedings	NNP
zkie-zero-2020	354	3	of	of	IN
zkie-zero-2020	354	4	EMNLP’11	EMNLP’11	NNP
zkie-zero-2020	354	5	,	,	,
zkie-zero-2020	354	6	pages	page	VBZ
zkie-zero-2020	354	7	782–792	782–792	CD
zkie-zero-2020	354	8	.	.	.
zkie-zero-2020	355	1	Filip	Filip	NNP
zkie-zero-2020	355	2	Ilievski	Ilievski	NNP
zkie-zero-2020	355	3	,	,	,
zkie-zero-2020	355	4	Piek	Piek	NNP
zkie-zero-2020	355	5	Vossen	Vossen	NNP
zkie-zero-2020	355	6	,	,	,
zkie-zero-2020	355	7	and	and	CC
zkie-zero-2020	355	8	Stefan	Stefan	NNP
zkie-zero-2020	355	9	Schlobach	Schlobach	NNP
zkie-zero-2020	355	10	.	.	.
zkie-zero-2020	356	1	2018	2018	CD
zkie-zero-2020	356	2	.	.	.
zkie-zero-2020	357	1	Systematic	Systematic	NNP
zkie-zero-2020	357	2	Study	Study	NNP
zkie-zero-2020	357	3	of	of	IN
zkie-zero-2020	357	4	Long	Long	NNP
zkie-zero-2020	357	5	Tail	Tail	NNP
zkie-zero-2020	357	6	Phenomena	Phenomena	NNP
zkie-zero-2020	357	7	in	in	IN
zkie-zero-2020	357	8	En-	en-	JJ
zkie-zero-2020	357	9	tity	tity	NN
zkie-zero-2020	357	10	Linking	linking	NN
zkie-zero-2020	357	11	.	.	.
zkie-zero-2020	358	1	In	in	IN
zkie-zero-2020	358	2	Proceedings	proceeding	NNS
zkie-zero-2020	358	3	of	of	IN
zkie-zero-2020	358	4	the	the	DT
zkie-zero-2020	358	5	27th	27th	NN
zkie-zero-2020	358	6	Inter-	Inter-	NNP
zkie-zero-2020	358	7	national	national	JJ
zkie-zero-2020	358	8	Conference	Conference	NNP
zkie-zero-2020	358	9	on	on	IN
zkie-zero-2020	358	10	Computational	Computational	NNP
zkie-zero-2020	358	11	Linguistics	Linguistics	NNP
zkie-zero-2020	358	12	,	,	,
zkie-zero-2020	358	13	pages	page	VBZ
zkie-zero-2020	358	14	664–674	664–674	CD
zkie-zero-2020	358	15	.	.	.
zkie-zero-2020	359	1	Thorsten	Thorsten	NNP
zkie-zero-2020	359	2	Joachims	joachim	NNS
zkie-zero-2020	359	3	.	.	.
zkie-zero-2020	360	1	2002	2002	CD
zkie-zero-2020	360	2	.	.	.
zkie-zero-2020	361	1	Optimizing	optimize	VBG
zkie-zero-2020	361	2	search	search	NN
zkie-zero-2020	361	3	engines	engine	NNS
zkie-zero-2020	361	4	using	use	VBG
zkie-zero-2020	361	5	clickthrough	clickthrough	NNP
zkie-zero-2020	361	6	data	datum	NNS
zkie-zero-2020	361	7	.	.	.
zkie-zero-2020	362	1	In	in	IN
zkie-zero-2020	362	2	Proceedings	proceeding	NNS
zkie-zero-2020	362	3	of	of	IN
zkie-zero-2020	362	4	the	the	DT
zkie-zero-2020	362	5	eighth	eighth	JJ
zkie-zero-2020	362	6	ACM	ACM	NNP
zkie-zero-2020	362	7	SIGKDD	SIGKDD	NNP
zkie-zero-2020	362	8	international	international	JJ
zkie-zero-2020	362	9	conference	conference	NN
zkie-zero-2020	362	10	on	on	IN
zkie-zero-2020	362	11	Knowledge	Knowledge	NNP
zkie-zero-2020	362	12	discovery	discovery	NN
zkie-zero-2020	362	13	and	and	CC
zkie-zero-2020	362	14	data	datum	NNS
zkie-zero-2020	362	15	mining	mining	NN
zkie-zero-2020	362	16	-	-	:
zkie-zero-2020	362	17	KDD	KDD	NNP
zkie-zero-2020	362	18	’	'	''
zkie-zero-2020	362	19	02	02	CD
zkie-zero-2020	362	20	,	,	,
zkie-zero-2020	362	21	pages	page	VBZ
zkie-zero-2020	362	22	133–142	133–142	CD
zkie-zero-2020	362	23	.	.	.
zkie-zero-2020	363	1	Guolin	Guolin	NNP
zkie-zero-2020	363	2	Ke	Ke	NNP
zkie-zero-2020	363	3	,	,	,
zkie-zero-2020	363	4	Qi	Qi	NNP
zkie-zero-2020	363	5	Meng	Meng	NNP
zkie-zero-2020	363	6	,	,	,
zkie-zero-2020	363	7	Thomas	Thomas	NNP
zkie-zero-2020	363	8	Finley	Finley	NNP
zkie-zero-2020	363	9	,	,	,
zkie-zero-2020	363	10	Taifeng	Taifeng	NNP
zkie-zero-2020	363	11	Wang	Wang	NNP
zkie-zero-2020	363	12	,	,	,
zkie-zero-2020	363	13	Wei	Wei	NNP
zkie-zero-2020	363	14	Chen	Chen	NNP
zkie-zero-2020	363	15	,	,	,
zkie-zero-2020	363	16	Weidong	Weidong	NNP
zkie-zero-2020	363	17	Ma	Ma	NNP
zkie-zero-2020	363	18	,	,	,
zkie-zero-2020	363	19	Qiwei	Qiwei	NNP
zkie-zero-2020	363	20	Ye	Ye	NNP
zkie-zero-2020	363	21	,	,	,
zkie-zero-2020	363	22	and	and	CC
zkie-zero-2020	363	23	Tie	Tie	NNP
zkie-zero-2020	363	24	-	-	HYPH
zkie-zero-2020	363	25	Yan	Yan	NNP
zkie-zero-2020	363	26	Liu	Liu	NNP
zkie-zero-2020	363	27	.	.	.
zkie-zero-2020	364	1	2017	2017	CD
zkie-zero-2020	364	2	.	.	.
zkie-zero-2020	365	1	LightGBM	LightGBM	NNP
zkie-zero-2020	365	2	:	:	:
zkie-zero-2020	365	3	A	a	DT
zkie-zero-2020	365	4	Highly	highly	RB
zkie-zero-2020	365	5	Efficient	efficient	JJ
zkie-zero-2020	365	6	Gradient	Gradient	NNP
zkie-zero-2020	365	7	Boosting	Boosting	NNP
zkie-zero-2020	365	8	Decision	Decision	NNP
zkie-zero-2020	365	9	Tree	Tree	NNP
zkie-zero-2020	365	10	.	.	.
zkie-zero-2020	366	1	In	in	IN
zkie-zero-2020	366	2	I.	I.	NNP
zkie-zero-2020	366	3	Guyon	Guyon	NNP
zkie-zero-2020	366	4	,	,	,
zkie-zero-2020	366	5	U.	U.	NNP
zkie-zero-2020	366	6	V.	V.	NNP
zkie-zero-2020	366	7	Luxburg	Luxburg	NNP
zkie-zero-2020	366	8	,	,	,
zkie-zero-2020	366	9	S.	S.	NNP
zkie-zero-2020	366	10	Bengio	Bengio	NNP
zkie-zero-2020	366	11	,	,	,
zkie-zero-2020	366	12	H.	H.	NNP
zkie-zero-2020	366	13	Wallach	Wallach	NNP
zkie-zero-2020	366	14	,	,	,
zkie-zero-2020	366	15	R.	R.	NNP
zkie-zero-2020	366	16	Fergus	Fergus	NNP
zkie-zero-2020	366	17	,	,	,
zkie-zero-2020	366	18	S.	S.	NNP
zkie-zero-2020	366	19	Vishwanathan	Vishwanathan	NNP
zkie-zero-2020	366	20	,	,	,
zkie-zero-2020	366	21	and	and	CC
zkie-zero-2020	366	22	R.	R.	NNP
zkie-zero-2020	366	23	Garnett	Garnett	NNP
zkie-zero-2020	366	24	,	,	,
zkie-zero-2020	366	25	editors	editor	NNS
zkie-zero-2020	366	26	,	,	,
zkie-zero-2020	366	27	Advances	Advances	NNPS
zkie-zero-2020	366	28	in	in	IN
zkie-zero-2020	366	29	Neural	Neural	NNP
zkie-zero-2020	366	30	Informa-	Informa-	NNP
zkie-zero-2020	366	31	tion	tion	NN
zkie-zero-2020	366	32	Processing	Processing	NNP
zkie-zero-2020	366	33	Systems	Systems	NNP
zkie-zero-2020	366	34	30	30	CD
zkie-zero-2020	366	35	,	,	,
zkie-zero-2020	366	36	pages	page	VBZ
zkie-zero-2020	366	37	3146–3154	3146–3154	CD
zkie-zero-2020	366	38	.	.	.
zkie-zero-2020	367	1	Jan	Jan	NNP
zkie-zero-2020	367	2	-	-	HYPH
zkie-zero-2020	367	3	Christoph	Christoph	NNP
zkie-zero-2020	367	4	Klie	Klie	NNP
zkie-zero-2020	367	5	,	,	,
zkie-zero-2020	367	6	Michael	Michael	NNP
zkie-zero-2020	367	7	Bugert	Bugert	NNP
zkie-zero-2020	367	8	,	,	,
zkie-zero-2020	367	9	Beto	Beto	NNP
zkie-zero-2020	367	10	Boullosa	Boullosa	NNP
zkie-zero-2020	367	11	,	,	,
zkie-zero-2020	367	12	Richard	Richard	NNP
zkie-zero-2020	367	13	Eckart	Eckart	NNP
zkie-zero-2020	367	14	de	de	NNP
zkie-zero-2020	367	15	Castilho	Castilho	NNP
zkie-zero-2020	367	16	,	,	,
zkie-zero-2020	367	17	and	and	CC
zkie-zero-2020	367	18	Iryna	Iryna	NNP
zkie-zero-2020	367	19	Gurevych	Gurevych	NNP
zkie-zero-2020	367	20	.	.	.
zkie-zero-2020	368	1	2018	2018	CD
zkie-zero-2020	368	2	.	.	.
zkie-zero-2020	369	1	The	the	DT
zkie-zero-2020	369	2	INCEpTION	INCEpTION	NNP
zkie-zero-2020	369	3	Platform	platform	NN
zkie-zero-2020	369	4	:	:	:
zkie-zero-2020	369	5	Machine-	Machine-	NNP
zkie-zero-2020	369	6	Assisted	Assisted	NNP
zkie-zero-2020	369	7	and	and	CC
zkie-zero-2020	369	8	Knowledge	Knowledge	NNP
zkie-zero-2020	369	9	-	-	HYPH
zkie-zero-2020	369	10	Oriented	Oriented	NNP
zkie-zero-2020	369	11	Interactive	Interactive	NNP
zkie-zero-2020	369	12	Anno-	Anno-	NNP
zkie-zero-2020	369	13	tation	tation	NN
zkie-zero-2020	369	14	.	.	.
zkie-zero-2020	370	1	In	in	IN
zkie-zero-2020	370	2	Proceedings	proceeding	NNS
zkie-zero-2020	370	3	of	of	IN
zkie-zero-2020	370	4	the	the	DT
zkie-zero-2020	370	5	27th	27th	JJ
zkie-zero-2020	370	6	International	International	NNP
zkie-zero-2020	370	7	Conference	Conference	NNP
zkie-zero-2020	370	8	on	on	IN
zkie-zero-2020	370	9	Computational	Computational	NNP
zkie-zero-2020	370	10	Linguistics	Linguistics	NNPS
zkie-zero-2020	370	11	:	:	:
zkie-zero-2020	370	12	System	System	NNP
zkie-zero-2020	370	13	Demonstrations	Demonstrations	NNPS
zkie-zero-2020	370	14	,	,	,
zkie-zero-2020	370	15	pages	page	NNS
zkie-zero-2020	370	16	5–9	5–9	CD
zkie-zero-2020	370	17	.	.	.
zkie-zero-2020	371	1	Edgar	Edgar	NNP
zkie-zero-2020	371	2	Meij	Meij	NNP
zkie-zero-2020	371	3	,	,	,
zkie-zero-2020	371	4	Krisztian	Krisztian	NNP
zkie-zero-2020	371	5	Balog	Balog	NNP
zkie-zero-2020	371	6	,	,	,
zkie-zero-2020	371	7	and	and	CC
zkie-zero-2020	371	8	Daan	Daan	NNP
zkie-zero-2020	371	9	Odijk	Odijk	NNP
zkie-zero-2020	371	10	.	.	.
zkie-zero-2020	372	1	2014	2014	CD
zkie-zero-2020	372	2	.	.	.
zkie-zero-2020	373	1	Entity	entity	NN
zkie-zero-2020	373	2	linking	linking	NN
zkie-zero-2020	373	3	and	and	CC
zkie-zero-2020	373	4	retrieval	retrieval	NN
zkie-zero-2020	373	5	for	for	IN
zkie-zero-2020	373	6	semantic	semantic	JJ
zkie-zero-2020	373	7	search	search	NN
zkie-zero-2020	373	8	.	.	.
zkie-zero-2020	374	1	In	in	IN
zkie-zero-2020	374	2	Proceedings	proceeding	NNS
zkie-zero-2020	374	3	of	of	IN
zkie-zero-2020	374	4	the	the	DT
zkie-zero-2020	374	5	7th	7th	JJ
zkie-zero-2020	374	6	ACM	ACM	NNP
zkie-zero-2020	374	7	international	international	JJ
zkie-zero-2020	374	8	confer-	confer-	JJ
zkie-zero-2020	374	9	ence	ence	NN
zkie-zero-2020	374	10	on	on	IN
zkie-zero-2020	374	11	Web	web	NN
zkie-zero-2020	374	12	search	search	NN
zkie-zero-2020	374	13	and	and	CC
zkie-zero-2020	374	14	data	datum	NNS
zkie-zero-2020	374	15	mining	mining	NN
zkie-zero-2020	374	16	-	-	:
zkie-zero-2020	374	17	WSDM	WSDM	NNP
zkie-zero-2020	374	18	'	'	POS
zkie-zero-2020	374	19	14	14	CD
zkie-zero-2020	374	20	,	,	,
zkie-zero-2020	374	21	pages	page	VBZ
zkie-zero-2020	374	22	683–684	683–684	CD
zkie-zero-2020	374	23	.	.	.
zkie-zero-2020	375	1	John	John	NNP
zkie-zero-2020	375	2	Melson	Melson	NNP
zkie-zero-2020	375	3	and	and	CC
zkie-zero-2020	375	4	Julia	Julia	NNP
zkie-zero-2020	375	5	Flanders	Flanders	NNPS
zkie-zero-2020	375	6	.	.	.
zkie-zero-2020	376	1	2010	2010	CD
zkie-zero-2020	376	2	.	.	.
zkie-zero-2020	377	1	Not	not	RB
zkie-zero-2020	377	2	Just	just	RB
zkie-zero-2020	377	3	One	one	CD
zkie-zero-2020	377	4	of	of	IN
zkie-zero-2020	377	5	Your	-PRON-	PRP$
zkie-zero-2020	377	6	Holiday	Holiday	NNP
zkie-zero-2020	377	7	Games	Games	NNPS
zkie-zero-2020	377	8	:	:	:
zkie-zero-2020	377	9	Names	name	NNS
zkie-zero-2020	377	10	and	and	CC
zkie-zero-2020	377	11	Name	name	VB
zkie-zero-2020	377	12	Encoding	encode	VBG
zkie-zero-2020	377	13	in	in	IN
zkie-zero-2020	377	14	the	the	DT
zkie-zero-2020	377	15	Women	Women	NNPS
zkie-zero-2020	377	16	Writers	Writers	NNPS
zkie-zero-2020	377	17	Project	Project	NNP
zkie-zero-2020	377	18	Textbase	Textbase	NNP
zkie-zero-2020	377	19	.	.	.
zkie-zero-2020	378	1	White	white	JJ
zkie-zero-2020	378	2	paper	paper	NN
zkie-zero-2020	378	3	,	,	,
zkie-zero-2020	378	4	Women	Women	NNPS
zkie-zero-2020	378	5	Writers	Writers	NNPS
zkie-zero-2020	378	6	Project	Project	NNP
zkie-zero-2020	378	7	,	,	,
zkie-zero-2020	378	8	Brown	Brown	NNP
zkie-zero-2020	378	9	University	University	NNP
zkie-zero-2020	378	10	.	.	.
zkie-zero-2020	379	1	David	David	NNP
zkie-zero-2020	379	2	Milne	Milne	NNP
zkie-zero-2020	379	3	and	and	CC
zkie-zero-2020	379	4	Ian	Ian	NNP
zkie-zero-2020	379	5	H.	H.	NNP
zkie-zero-2020	379	6	Witten	Witten	NNP
zkie-zero-2020	379	7	.	.	.
zkie-zero-2020	380	1	2008	2008	CD
zkie-zero-2020	380	2	.	.	.
zkie-zero-2020	381	1	Learning	learn	VBG
zkie-zero-2020	381	2	to	to	TO
zkie-zero-2020	381	3	link	link	VB
zkie-zero-2020	381	4	with	with	IN
zkie-zero-2020	381	5	Wikipedia	Wikipedia	NNP
zkie-zero-2020	381	6	.	.	.
zkie-zero-2020	382	1	In	in	IN
zkie-zero-2020	382	2	Proceeding	Proceeding	NNP
zkie-zero-2020	382	3	of	of	IN
zkie-zero-2020	382	4	the	the	DT
zkie-zero-2020	382	5	17th	17th	JJ
zkie-zero-2020	382	6	ACM	ACM	NNP
zkie-zero-2020	382	7	conference	conference	NN
zkie-zero-2020	382	8	on	on	IN
zkie-zero-2020	382	9	Information	information	NN
zkie-zero-2020	382	10	and	and	CC
zkie-zero-2020	382	11	knowledge	knowledge	NN
zkie-zero-2020	382	12	mining	mining	NN
zkie-zero-2020	382	13	-	-	:
zkie-zero-2020	382	14	CIKM	CIKM	NNP
zkie-zero-2020	382	15	'	'	POS
zkie-zero-2020	382	16	08	08	CD
zkie-zero-2020	382	17	,	,	,
zkie-zero-2020	382	18	pages	page	NNS
zkie-zero-2020	382	19	509–518	509–518	CD
zkie-zero-2020	382	20	.	.	.
zkie-zero-2020	383	1	Gary	Gary	NNP
zkie-zero-2020	383	2	Munnelly	Munnelly	NNP
zkie-zero-2020	383	3	and	and	CC
zkie-zero-2020	383	4	Séamus	séamus	VB
zkie-zero-2020	383	5	Lawless	Lawless	NNP
zkie-zero-2020	383	6	.	.	.
zkie-zero-2020	384	1	2018a	2018a	CD
zkie-zero-2020	384	2	.	.	.
zkie-zero-2020	385	1	Con-	Con-	NNP
zkie-zero-2020	385	2	structing	structe	VBG
zkie-zero-2020	385	3	a	a	DT
zkie-zero-2020	385	4	knowledge	knowledge	NN
zkie-zero-2020	385	5	base	base	NN
zkie-zero-2020	385	6	for	for	IN
zkie-zero-2020	385	7	entity	entity	NN
zkie-zero-2020	385	8	linking	link	VBG
zkie-zero-2020	385	9	on	on	IN
zkie-zero-2020	385	10	Irish	irish	JJ
zkie-zero-2020	385	11	cultural	cultural	JJ
zkie-zero-2020	385	12	heritage	heritage	NN
zkie-zero-2020	385	13	collections	collection	NNS
zkie-zero-2020	385	14	.	.	.
zkie-zero-2020	386	1	Procedia	Procedia	NNP
zkie-zero-2020	386	2	Computer	Computer	NNP
zkie-zero-2020	386	3	Science	Science	NNP
zkie-zero-2020	386	4	,	,	,
zkie-zero-2020	386	5	137:199–210	137:199–210	CD
zkie-zero-2020	386	6	.	.	.
zkie-zero-2020	387	1	Gary	Gary	NNP
zkie-zero-2020	387	2	Munnelly	Munnelly	NNP
zkie-zero-2020	387	3	and	and	CC
zkie-zero-2020	387	4	Seamus	Seamus	NNP
zkie-zero-2020	387	5	Lawless	Lawless	NNP
zkie-zero-2020	387	6	.	.	.
zkie-zero-2020	388	1	2018b	2018b	LS
zkie-zero-2020	388	2	.	.	.
zkie-zero-2020	389	1	Investi-	Investi-	NNP
zkie-zero-2020	389	2	gating	gate	VBG
zkie-zero-2020	389	3	Entity	Entity	NNP
zkie-zero-2020	389	4	Linking	link	VBG
zkie-zero-2020	389	5	in	in	IN
zkie-zero-2020	389	6	Early	early	JJ
zkie-zero-2020	389	7	English	English	NNP
zkie-zero-2020	389	8	Legal	Legal	NNP
zkie-zero-2020	389	9	Doc-	Doc-	NNP
zkie-zero-2020	389	10	uments	ument	NNS
zkie-zero-2020	389	11	.	.	.
zkie-zero-2020	390	1	In	in	IN
zkie-zero-2020	390	2	Proceedings	proceeding	NNS
zkie-zero-2020	390	3	of	of	IN
zkie-zero-2020	390	4	the	the	DT
zkie-zero-2020	390	5	18th	18th	JJ
zkie-zero-2020	390	6	ACM	ACM	NNP
zkie-zero-2020	390	7	/	/	,
zkie-zero-2020	390	8	IEEE	IEEE	NNP
zkie-zero-2020	390	9	on	on	IN
zkie-zero-2020	390	10	Joint	Joint	NNP
zkie-zero-2020	390	11	Conference	Conference	NNP
zkie-zero-2020	390	12	on	on	IN
zkie-zero-2020	390	13	Digital	Digital	NNP
zkie-zero-2020	390	14	Libraries	Libraries	NNPS
zkie-zero-2020	390	15	-	-	:
zkie-zero-2020	390	16	JCDL	JCDL	NNP
zkie-zero-2020	390	17	’	'	''
zkie-zero-2020	390	18	18	18	CD
zkie-zero-2020	390	19	,	,	,
zkie-zero-2020	390	20	pages	page	NNS
zkie-zero-2020	390	21	59–68	59–68	CD
zkie-zero-2020	390	22	.	.	.
zkie-zero-2020	391	1	Farhad	Farhad	NNP
zkie-zero-2020	391	2	Nooralahzadeh	Nooralahzadeh	NNP
zkie-zero-2020	391	3	and	and	CC
zkie-zero-2020	391	4	Lilja	Lilja	NNP
zkie-zero-2020	391	5	Øvrelid	Øvrelid	NNP
zkie-zero-2020	391	6	.	.	.
zkie-zero-2020	392	1	2018	2018	CD
zkie-zero-2020	392	2	.	.	.
zkie-zero-2020	393	1	SIRIUS	SIRIUS	NNP
zkie-zero-2020	393	2	-	-	HYPH
zkie-zero-2020	393	3	LTG	LTG	NNP
zkie-zero-2020	393	4	:	:	:
zkie-zero-2020	393	5	An	an	DT
zkie-zero-2020	393	6	Entity	Entity	NNP
zkie-zero-2020	393	7	Linking	link	VBG
zkie-zero-2020	393	8	Approach	approach	NN
zkie-zero-2020	393	9	to	to	IN
zkie-zero-2020	393	10	Fact	Fact	NNP
zkie-zero-2020	393	11	Extraction	Extraction	NNP
zkie-zero-2020	393	12	and	and	CC
zkie-zero-2020	393	13	Verification	Verification	NNP
zkie-zero-2020	393	14	.	.	.
zkie-zero-2020	394	1	In	in	IN
zkie-zero-2020	394	2	Proceedings	Proceedings	NNP
zkie-zero-2020	394	3	of	of	IN
zkie-zero-2020	394	4	the	the	DT
zkie-zero-2020	394	5	First	First	NNP
zkie-zero-2020	394	6	Workshop	Workshop	NNP
zkie-zero-2020	394	7	on	on	IN
zkie-zero-2020	394	8	Fact	Fact	NNP
zkie-zero-2020	394	9	Extraction	Extraction	NNP
zkie-zero-2020	394	10	and	and	CC
zkie-zero-2020	394	11	VERification	verification	NN
zkie-zero-2020	394	12	(	(	-LRB-
zkie-zero-2020	394	13	FEVER	FEVER	NNP
zkie-zero-2020	394	14	)	)	-RRB-
zkie-zero-2020	394	15	,	,	,
zkie-zero-2020	394	16	pages	page	VBZ
zkie-zero-2020	394	17	119–123	119–123	CD
zkie-zero-2020	394	18	.	.	.
zkie-zero-2020	395	1	Nils	Nils	NNP
zkie-zero-2020	395	2	Reimers	Reimers	NNPS
zkie-zero-2020	395	3	and	and	CC
zkie-zero-2020	395	4	Iryna	Iryna	NNP
zkie-zero-2020	395	5	Gurevych	Gurevych	NNP
zkie-zero-2020	395	6	.	.	.
zkie-zero-2020	396	1	2019	2019	CD
zkie-zero-2020	396	2	.	.	.
zkie-zero-2020	397	1	Sentence-	Sentence-	NNP
zkie-zero-2020	397	2	BERT	BERT	NNP
zkie-zero-2020	397	3	:	:	:
zkie-zero-2020	397	4	Sentence	Sentence	NNP
zkie-zero-2020	397	5	Embeddings	embedding	NNS
zkie-zero-2020	397	6	using	use	VBG
zkie-zero-2020	397	7	Siamese	Siamese	NNP
zkie-zero-2020	397	8	BERT-	BERT-	NNP
zkie-zero-2020	397	9	Networks	Networks	NNP
zkie-zero-2020	397	10	.	.	.
zkie-zero-2020	398	1	In	in	IN
zkie-zero-2020	398	2	Proceedings	Proceedings	NNP
zkie-zero-2020	398	3	of	of	IN
zkie-zero-2020	398	4	the	the	DT
zkie-zero-2020	398	5	2019	2019	CD
zkie-zero-2020	398	6	Conference	Conference	NNP
zkie-zero-2020	398	7	on	on	IN
zkie-zero-2020	398	8	Empirical	Empirical	NNP
zkie-zero-2020	398	9	Methods	Methods	NNPS
zkie-zero-2020	398	10	in	in	IN
zkie-zero-2020	398	11	Natural	Natural	NNP
zkie-zero-2020	398	12	Language	Language	NNP
zkie-zero-2020	398	13	Process-	Process-	NNP
zkie-zero-2020	398	14	ing	ing	NNP
zkie-zero-2020	398	15	,	,	,
zkie-zero-2020	398	16	pages	page	NNS
zkie-zero-2020	398	17	3980–3990	3980–3990	CD
zkie-zero-2020	398	18	.	.	.
zkie-zero-2020	399	1	Matthias	Matthias	NNP
zkie-zero-2020	399	2	Schlögl	schlögl	ADD
zkie-zero-2020	399	3	and	and	CC
zkie-zero-2020	399	4	Katalin	Katalin	NNP
zkie-zero-2020	399	5	Lejtovicz	Lejtovicz	NNP
zkie-zero-2020	399	6	.	.	.
zkie-zero-2020	400	1	2017	2017	CD
zkie-zero-2020	400	2	.	.	.
zkie-zero-2020	401	1	APIS	APIS	NNP
zkie-zero-2020	401	2	-	-	HYPH
zkie-zero-2020	401	3	Austrian	austrian	JJ
zkie-zero-2020	401	4	Prosopographical	Prosopographical	NNP
zkie-zero-2020	401	5	Information	Information	NNP
zkie-zero-2020	401	6	System	System	NNP
zkie-zero-2020	401	7	.	.	.
zkie-zero-2020	402	1	In	in	IN
zkie-zero-2020	402	2	Proceedings	proceeding	NNS
zkie-zero-2020	402	3	of	of	IN
zkie-zero-2020	402	4	the	the	DT
zkie-zero-2020	402	5	Second	Second	NNP
zkie-zero-2020	402	6	Conference	Conference	NNP
zkie-zero-2020	402	7	on	on	IN
zkie-zero-2020	402	8	Biograph-	Biograph-	NNP
zkie-zero-2020	402	9	ical	ical	NNP
zkie-zero-2020	402	10	Data	Data	NNP
zkie-zero-2020	402	11	in	in	IN
zkie-zero-2020	402	12	a	a	DT
zkie-zero-2020	402	13	Digital	Digital	NNP
zkie-zero-2020	402	14	World	World	NNP
zkie-zero-2020	402	15	2017	2017	CD
zkie-zero-2020	402	16	.	.	.
zkie-zero-2020	402	17	https://doi.org/10.1162/jmlr.2003.3.4-5.951	https://doi.org/10.1162/jmlr.2003.3.4-5.951	NNP
zkie-zero-2020	402	18	https://doi.org/10.1162/jmlr.2003.3.4-5.951	https://doi.org/10.1162/jmlr.2003.3.4-5.951	NNP
zkie-zero-2020	402	19	https://doi.org/10.1145/2506182.2506198	https://doi.org/10.1145/2506182.2506198	ADD
zkie-zero-2020	402	20	https://doi.org/10.1145/2506182.2506198	https://doi.org/10.1145/2506182.2506198	ADD
zkie-zero-2020	402	21	https://doi.org/10.1007/978-0-387-32833-1_169	https://doi.org/10.1007/978-0-387-32833-1_169	NNS
zkie-zero-2020	402	22	https://doi.org/10.1007/978-0-387-32833-1_169	https://doi.org/10.1007/978-0-387-32833-1_169	JJ
zkie-zero-2020	402	23	https://doi.org/10.18653/v1/n19-1231	https://doi.org/10.18653/v1/n19-1231	NN
zkie-zero-2020	402	24	https://doi.org/10.18653/v1/n19-1231	https://doi.org/10.18653/v1/n19-1231	NNS
zkie-zero-2020	402	25	https://doi.org/10.18653/v1/n19-1231	https://doi.org/10.18653/v1/n19-1231	NNS
zkie-zero-2020	402	26	https://doi.org/10.1145/1871437.1871689	https://doi.org/10.1145/1871437.1871689	ADD
zkie-zero-2020	402	27	https://doi.org/10.1145/1871437.1871689	https://doi.org/10.1145/1871437.1871689	ADD
zkie-zero-2020	402	28	https://doi.org/10.1145/1871437.1871689	https://doi.org/10.1145/1871437.1871689	ADD
zkie-zero-2020	402	29	https://doi.org/10.18653/v1/d18-1445	https://doi.org/10.18653/v1/d18-1445	NNP
zkie-zero-2020	402	30	https://doi.org/10.18653/v1/d18-1445	https://doi.org/10.18653/v1/d18-1445	NNP
zkie-zero-2020	402	31	https://doi.org/10.18653/v1/d18-1445	https://doi.org/10.18653/v1/d18-1445	NNP
zkie-zero-2020	402	32	https://www.aclweb.org/anthology/N13-1122	https://www.aclweb.org/anthology/n13-1122	UH
zkie-zero-2020	402	33	https://www.aclweb.org/anthology/N13-1122	https://www.aclweb.org/anthology/N13-1122	NNP
zkie-zero-2020	402	34	https://doi.org/10.1023/a:1012487302797	https://doi.org/10.1023/a:1012487302797	NNP
zkie-zero-2020	402	35	https://doi.org/10.1023/a:1012487302797	https://doi.org/10.1023/a:1012487302797	NNP
zkie-zero-2020	402	36	https://doi.org/10.18653/v1/d16-1258	https://doi.org/10.18653/v1/d16-1258	NNP
zkie-zero-2020	402	37	https://www.aclweb.org/anthology/D11-1072.pdf	https://www.aclweb.org/anthology/D11-1072.pdf	NNP
zkie-zero-2020	402	38	https://www.aclweb.org/anthology/D11-1072.pdf	https://www.aclweb.org/anthology/D11-1072.pdf	NNP
zkie-zero-2020	402	39	https://www.aclweb.org/anthology/C18-1056	https://www.aclweb.org/anthology/c18-1056	ADD
zkie-zero-2020	402	40	https://www.aclweb.org/anthology/C18-1056	https://www.aclweb.org/anthology/c18-1056	ADD
zkie-zero-2020	402	41	https://doi.org/10.1145/775047.775067	https://doi.org/10.1145/775047.775067	ADD
zkie-zero-2020	402	42	https://doi.org/10.1145/775047.775067	https://doi.org/10.1145/775047.775067	ADD
zkie-zero-2020	402	43	http://papers.nips.cc/paper/6907-lightgbm-a-highly-efficient-gradient-boosting-decision-tree.pdf	http://papers.nips.cc/paper/6907-lightgbm-a-highly-efficient-gradient-boosting-decision-tree.pdf	ADD
zkie-zero-2020	402	44	http://papers.nips.cc/paper/6907-lightgbm-a-highly-efficient-gradient-boosting-decision-tree.pdf	http://papers.nips.cc/paper/6907-lightgbm-a-highly-efficient-gradient-boosting-decision-tree.pdf	ADD
zkie-zero-2020	402	45	https://www.aclweb.org/anthology/C18-2002	https://www.aclweb.org/anthology/c18-2002	UH
zkie-zero-2020	402	46	https://www.aclweb.org/anthology/C18-2002	https://www.aclweb.org/anthology/c18-2002	ADD
zkie-zero-2020	402	47	https://www.aclweb.org/anthology/C18-2002	https://www.aclweb.org/anthology/c18-2002	ADD
zkie-zero-2020	402	48	https://doi.org/10.1145/2556195.2556201	https://doi.org/10.1145/2556195.2556201	IN
zkie-zero-2020	402	49	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	NNP
zkie-zero-2020	402	50	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	NNP
zkie-zero-2020	402	51	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	https://www.wwp.northeastern.edu/research/publications/reports/neh_2008/WWP_Names_White_Paper.pdf	NNP
zkie-zero-2020	402	52	https://doi.org/10.1145/1458082.1458150	https://doi.org/10.1145/1458082.1458150	FW
zkie-zero-2020	402	53	https://doi.org/10.1145/1458082.1458150	https://doi.org/10.1145/1458082.1458150	FW
zkie-zero-2020	402	54	https://doi.org/10.1016/j.procs.2018.09.019	https://doi.org/10.1016/j.procs.2018.09.019	FW
zkie-zero-2020	402	55	https://doi.org/10.1016/j.procs.2018.09.019	https://doi.org/10.1016/j.procs.2018.09.019	FW
zkie-zero-2020	402	56	https://doi.org/10.1016/j.procs.2018.09.019	https://doi.org/10.1016/j.procs.2018.09.019	XX
zkie-zero-2020	402	57	https://doi.org/10.1145/3197026.3197055	https://doi.org/10.1145/3197026.3197055	ADD
zkie-zero-2020	402	58	https://doi.org/10.1145/3197026.3197055	https://doi.org/10.1145/3197026.3197055	UH
zkie-zero-2020	402	59	https://doi.org/10.1145/3197026.3197055	https://doi.org/10.1145/3197026.3197055	NFP
zkie-zero-2020	402	60	https://doi.org/10.18653/v1/w18-5519	https://doi.org/10.18653/v1/w18-5519	NNP
zkie-zero-2020	402	61	https://doi.org/10.18653/v1/w18-5519	https://doi.org/10.18653/v1/w18-5519	NNP
zkie-zero-2020	402	62	https://doi.org/10.18653/v1/D19-1410	https://doi.org/10.18653/v1/D19-1410	NNP
zkie-zero-2020	402	63	https://doi.org/10.18653/v1/D19-1410	https://doi.org/10.18653/v1/D19-1410	NNP
zkie-zero-2020	402	64	https://doi.org/10.18653/v1/D19-1410	https://doi.org/10.18653/v1/D19-1410	NNP
zkie-zero-2020	402	65	https://doi.org/10.5281/zenodo.846571	https://doi.org/10.5281/zenodo.846571	NN
zkie-zero-2020	402	66	https://doi.org/10.5281/zenodo.846571	https://doi.org/10.5281/zenodo.846571	NN
zkie-zero-2020	402	67	6992	6992	CD
zkie-zero-2020	402	68	Klaus	Klaus	NNP
zkie-zero-2020	402	69	U.	U.	NNP
zkie-zero-2020	402	70	Schulz	Schulz	NNP
zkie-zero-2020	402	71	and	and	CC
zkie-zero-2020	402	72	Stoyan	Stoyan	NNP
zkie-zero-2020	402	73	Mihov	Mihov	NNP
zkie-zero-2020	402	74	.	.	.
zkie-zero-2020	403	1	2002	2002	CD
zkie-zero-2020	403	2	.	.	.
zkie-zero-2020	404	1	Fast	fast	JJ
zkie-zero-2020	404	2	string	string	NN
zkie-zero-2020	404	3	correction	correction	NN
zkie-zero-2020	404	4	with	with	IN
zkie-zero-2020	404	5	Levenshtein	Levenshtein	NNP
zkie-zero-2020	404	6	automata	automata	NN
zkie-zero-2020	404	7	.	.	.
zkie-zero-2020	405	1	Interna-	Interna-	NNP
zkie-zero-2020	405	2	tional	tional	JJ
zkie-zero-2020	405	3	Journal	Journal	NNP
zkie-zero-2020	405	4	on	on	IN
zkie-zero-2020	405	5	Document	Document	NNP
zkie-zero-2020	405	6	Analysis	Analysis	NNP
zkie-zero-2020	405	7	and	and	CC
zkie-zero-2020	405	8	Recogni-	Recogni-	NNP
zkie-zero-2020	405	9	tion	tion	NN
zkie-zero-2020	405	10	,	,	,
zkie-zero-2020	405	11	5(1):67–85	5(1):67–85	CD
zkie-zero-2020	405	12	.	.	.
zkie-zero-2020	406	1	Wei	Wei	NNP
zkie-zero-2020	406	2	Shen	Shen	NNP
zkie-zero-2020	406	3	,	,	,
zkie-zero-2020	406	4	Jianyong	Jianyong	NNP
zkie-zero-2020	406	5	Wang	Wang	NNP
zkie-zero-2020	406	6	,	,	,
zkie-zero-2020	406	7	and	and	CC
zkie-zero-2020	406	8	Jiawei	Jiawei	NNP
zkie-zero-2020	406	9	Han	Han	NNP
zkie-zero-2020	406	10	.	.	.
zkie-zero-2020	407	1	2015	2015	CD
zkie-zero-2020	407	2	.	.	.
zkie-zero-2020	408	1	En-	en-	CC
zkie-zero-2020	408	2	tity	tity	NN
zkie-zero-2020	408	3	Linking	link	VBG
zkie-zero-2020	408	4	with	with	IN
zkie-zero-2020	408	5	a	a	DT
zkie-zero-2020	408	6	Knowledge	Knowledge	NNP
zkie-zero-2020	408	7	Base	Base	NNP
zkie-zero-2020	408	8	:	:	:
zkie-zero-2020	408	9	Issues	issue	NNS
zkie-zero-2020	408	10	,	,	,
zkie-zero-2020	408	11	Tech-	tech-	NN
zkie-zero-2020	408	12	niques	nique	NNS
zkie-zero-2020	408	13	,	,	,
zkie-zero-2020	408	14	and	and	CC
zkie-zero-2020	408	15	Solutions	Solutions	NNPS
zkie-zero-2020	408	16	.	.	.
zkie-zero-2020	409	1	IEEE	IEEE	NNP
zkie-zero-2020	409	2	Transactions	Transactions	NNPS
zkie-zero-2020	409	3	on	on	IN
zkie-zero-2020	409	4	Knowl-	Knowl-	NNP
zkie-zero-2020	409	5	edge	edge	NN
zkie-zero-2020	409	6	and	and	CC
zkie-zero-2020	409	7	Data	Data	NNP
zkie-zero-2020	409	8	Engineering	Engineering	NNP
zkie-zero-2020	409	9	,	,	,
zkie-zero-2020	409	10	27(2):443–460	27(2):443–460	CD
zkie-zero-2020	409	11	.	.	.
zkie-zero-2020	410	1	Ricardo	Ricardo	NNP
zkie-zero-2020	410	2	Usbeck	Usbeck	NNP
zkie-zero-2020	410	3	,	,	,
zkie-zero-2020	410	4	Axel	Axel	NNP
zkie-zero-2020	410	5	-	-	HYPH
zkie-zero-2020	410	6	Cyrille	Cyrille	NNP
zkie-zero-2020	410	7	Ngonga	Ngonga	NNP
zkie-zero-2020	410	8	Ngomo	Ngomo	NNP
zkie-zero-2020	410	9	,	,	,
zkie-zero-2020	410	10	Michael	Michael	NNP
zkie-zero-2020	410	11	Röder	Röder	NNP
zkie-zero-2020	410	12	,	,	,
zkie-zero-2020	410	13	Daniel	Daniel	NNP
zkie-zero-2020	410	14	Gerber	Gerber	NNP
zkie-zero-2020	410	15	,	,	,
zkie-zero-2020	410	16	Sandro	Sandro	NNP
zkie-zero-2020	410	17	Athaide	Athaide	NNP
zkie-zero-2020	410	18	Coelho	Coelho	NNP
zkie-zero-2020	410	19	,	,	,
zkie-zero-2020	410	20	Sören	Sören	NNP
zkie-zero-2020	410	21	Auer	Auer	NNP
zkie-zero-2020	410	22	,	,	,
zkie-zero-2020	410	23	and	and	CC
zkie-zero-2020	410	24	Andreas	Andreas	NNP
zkie-zero-2020	410	25	Both	both	DT
zkie-zero-2020	410	26	.	.	.
zkie-zero-2020	411	1	2014	2014	CD
zkie-zero-2020	411	2	.	.	.
zkie-zero-2020	412	1	AGDISTIS	AGDISTIS	NNP
zkie-zero-2020	412	2	-	-	HYPH
zkie-zero-2020	412	3	Graph	Graph	NNP
zkie-zero-2020	412	4	-	-	HYPH
zkie-zero-2020	412	5	Based	base	VBN
zkie-zero-2020	412	6	Disambiguation	Disambiguation	NNP
zkie-zero-2020	412	7	of	of	IN
zkie-zero-2020	412	8	Named	Named	NNP
zkie-zero-2020	412	9	Entities	Entities	NNP
zkie-zero-2020	412	10	Using	use	VBG
zkie-zero-2020	412	11	Linked	Linked	NNP
zkie-zero-2020	412	12	Data	Data	NNP
zkie-zero-2020	412	13	.	.	.
zkie-zero-2020	413	1	In	in	IN
zkie-zero-2020	413	2	The	the	DT
zkie-zero-2020	413	3	Semantic	semantic	JJ
zkie-zero-2020	413	4	Web	web	NN
zkie-zero-2020	413	5	–	–	:
zkie-zero-2020	413	6	ISWC	ISWC	NNP
zkie-zero-2020	413	7	2014	2014	CD
zkie-zero-2020	413	8	,	,	,
zkie-zero-2020	413	9	pages	page	VBZ
zkie-zero-2020	413	10	457–471	457–471	CD
zkie-zero-2020	413	11	.	.	.
zkie-zero-2020	414	1	Eric	Eric	NNP
zkie-zero-2020	414	2	Wallace	Wallace	NNP
zkie-zero-2020	414	3	,	,	,
zkie-zero-2020	414	4	Pedro	Pedro	NNP
zkie-zero-2020	414	5	Rodriguez	Rodriguez	NNP
zkie-zero-2020	414	6	,	,	,
zkie-zero-2020	414	7	Shi	Shi	NNP
zkie-zero-2020	414	8	Feng	Feng	NNP
zkie-zero-2020	414	9	,	,	,
zkie-zero-2020	414	10	Ikuya	Ikuya	NNP
zkie-zero-2020	414	11	Ya-	Ya-	NNP
zkie-zero-2020	414	12	mada	mada	NN
zkie-zero-2020	414	13	,	,	,
zkie-zero-2020	414	14	and	and	CC
zkie-zero-2020	414	15	Jordan	Jordan	NNP
zkie-zero-2020	414	16	Boyd	Boyd	NNP
zkie-zero-2020	414	17	-	-	HYPH
zkie-zero-2020	414	18	Graber	Graber	NNP
zkie-zero-2020	414	19	.	.	.
zkie-zero-2020	415	1	2019	2019	CD
zkie-zero-2020	415	2	.	.	.
zkie-zero-2020	416	1	Trick	trick	VB
zkie-zero-2020	416	2	Me	-PRON-	PRP
zkie-zero-2020	416	3	If	if	IN
zkie-zero-2020	416	4	You	-PRON-	PRP
zkie-zero-2020	416	5	Can	Can	MD
zkie-zero-2020	416	6	:	:	:
zkie-zero-2020	416	7	Human	human	NN
zkie-zero-2020	416	8	-	-	HYPH
zkie-zero-2020	416	9	in	in	IN
zkie-zero-2020	416	10	-	-	HYPH
zkie-zero-2020	416	11	the	the	DT
zkie-zero-2020	416	12	-	-	HYPH
zkie-zero-2020	416	13	loop	loop	NN
zkie-zero-2020	416	14	Generation	Generation	NNP
zkie-zero-2020	416	15	of	of	IN
zkie-zero-2020	416	16	Ad-	Ad-	NNP
zkie-zero-2020	416	17	versarial	versarial	NNP
zkie-zero-2020	416	18	Question	Question	NNP
zkie-zero-2020	416	19	Answering	Answering	NNP
zkie-zero-2020	416	20	Examples	Examples	NNPS
zkie-zero-2020	416	21	.	.	.
zkie-zero-2020	417	1	Transac-	transac-	IN
zkie-zero-2020	417	2	tions	tion	NNS
zkie-zero-2020	417	3	of	of	IN
zkie-zero-2020	417	4	the	the	DT
zkie-zero-2020	417	5	Association	Association	NNP
zkie-zero-2020	417	6	for	for	IN
zkie-zero-2020	417	7	Computational	Computational	NNP
zkie-zero-2020	417	8	Linguis-	Linguis-	NNP
zkie-zero-2020	417	9	tics	tic	NNS
zkie-zero-2020	417	10	,	,	,
zkie-zero-2020	417	11	7(0):387–401	7(0):387–401	ADD
zkie-zero-2020	417	12	.	.	.
zkie-zero-2020	418	1	Travis	Travis	NNP
zkie-zero-2020	418	2	Wolfe	Wolfe	NNP
zkie-zero-2020	418	3	,	,	,
zkie-zero-2020	418	4	Mark	Mark	NNP
zkie-zero-2020	418	5	Dredze	Dredze	NNP
zkie-zero-2020	418	6	,	,	,
zkie-zero-2020	418	7	James	James	NNP
zkie-zero-2020	418	8	Mayfield	Mayfield	NNP
zkie-zero-2020	418	9	,	,	,
zkie-zero-2020	418	10	Paul	Paul	NNP
zkie-zero-2020	418	11	McNamee	McNamee	NNP
zkie-zero-2020	418	12	,	,	,
zkie-zero-2020	418	13	Craig	Craig	NNP
zkie-zero-2020	418	14	Harman	Harman	NNP
zkie-zero-2020	418	15	,	,	,
zkie-zero-2020	418	16	Tim	Tim	NNP
zkie-zero-2020	418	17	Finin	Finin	NNP
zkie-zero-2020	418	18	,	,	,
zkie-zero-2020	418	19	and	and	CC
zkie-zero-2020	418	20	Ben-	Ben-	NNP
zkie-zero-2020	418	21	jamin	jamin	NN
zkie-zero-2020	418	22	Van	Van	NNP
zkie-zero-2020	418	23	Durme	Durme	NNP
zkie-zero-2020	418	24	.	.	.
zkie-zero-2020	419	1	2015	2015	CD
zkie-zero-2020	419	2	.	.	.
zkie-zero-2020	420	1	Interactive	Interactive	NNP
zkie-zero-2020	420	2	Knowledge	Knowledge	NNP
zkie-zero-2020	420	3	Base	Base	NNP
zkie-zero-2020	420	4	Population	Population	NNP
zkie-zero-2020	420	5	.	.	.
zkie-zero-2020	421	1	Seid	Seid	NNP
zkie-zero-2020	421	2	Muhie	Muhie	NNP
zkie-zero-2020	421	3	Yimam	Yimam	NNP
zkie-zero-2020	421	4	,	,	,
zkie-zero-2020	421	5	Chris	Chris	NNP
zkie-zero-2020	421	6	Biemann	Biemann	NNP
zkie-zero-2020	421	7	,	,	,
zkie-zero-2020	421	8	Richard	Richard	NNP
zkie-zero-2020	421	9	Eckart	Eckart	NNP
zkie-zero-2020	421	10	de	de	NNP
zkie-zero-2020	421	11	Castilho	Castilho	NNP
zkie-zero-2020	421	12	,	,	,
zkie-zero-2020	421	13	and	and	CC
zkie-zero-2020	421	14	Iryna	Iryna	NNP
zkie-zero-2020	421	15	Gurevych	Gurevych	NNP
zkie-zero-2020	421	16	.	.	.
zkie-zero-2020	422	1	2014	2014	CD
zkie-zero-2020	422	2	.	.	.
zkie-zero-2020	423	1	Auto-	Auto-	NNP
zkie-zero-2020	423	2	matic	matic	JJ
zkie-zero-2020	423	3	Annotation	Annotation	NNP
zkie-zero-2020	423	4	Suggestions	Suggestions	NNP
zkie-zero-2020	423	5	and	and	CC
zkie-zero-2020	423	6	Custom	Custom	NNP
zkie-zero-2020	423	7	Annota-	Annota-	NNP
zkie-zero-2020	423	8	tion	tion	NN
zkie-zero-2020	423	9	Layers	Layers	NNPS
zkie-zero-2020	423	10	in	in	IN
zkie-zero-2020	423	11	WebAnno	WebAnno	NNP
zkie-zero-2020	423	12	.	.	.
zkie-zero-2020	424	1	In	in	IN
zkie-zero-2020	424	2	Proceedings	proceeding	NNS
zkie-zero-2020	424	3	of	of	IN
zkie-zero-2020	424	4	52nd	52nd	JJ
zkie-zero-2020	424	5	Annual	Annual	NNP
zkie-zero-2020	424	6	Meeting	Meeting	NNP
zkie-zero-2020	424	7	of	of	IN
zkie-zero-2020	424	8	the	the	DT
zkie-zero-2020	424	9	Association	Association	NNP
zkie-zero-2020	424	10	for	for	IN
zkie-zero-2020	424	11	Computa-	Computa-	NNP
zkie-zero-2020	424	12	tional	tional	JJ
zkie-zero-2020	424	13	Linguistics	Linguistics	NNPS
zkie-zero-2020	424	14	:	:	:
zkie-zero-2020	424	15	System	System	NNP
zkie-zero-2020	424	16	Demonstrations	Demonstrations	NNPS
zkie-zero-2020	424	17	,	,	,
zkie-zero-2020	424	18	pages	page	NNS
zkie-zero-2020	424	19	91–96	91–96	CD
zkie-zero-2020	424	20	.	.	.
zkie-zero-2020	425	1	Zhicheng	Zhicheng	NNP
zkie-zero-2020	425	2	Zheng	Zheng	NNP
zkie-zero-2020	425	3	,	,	,
zkie-zero-2020	425	4	Fangtao	Fangtao	NNP
zkie-zero-2020	425	5	Li	Li	NNP
zkie-zero-2020	425	6	,	,	,
zkie-zero-2020	425	7	Minlie	Minlie	NNP
zkie-zero-2020	425	8	Huang	Huang	NNP
zkie-zero-2020	425	9	,	,	,
zkie-zero-2020	425	10	and	and	CC
zkie-zero-2020	425	11	Xi-	xi-	VBP
zkie-zero-2020	425	12	aoyan	aoyan	JJ
zkie-zero-2020	425	13	Zhu	Zhu	NNP
zkie-zero-2020	425	14	.	.	.
zkie-zero-2020	426	1	2010	2010	CD
zkie-zero-2020	426	2	.	.	.
zkie-zero-2020	427	1	Learning	learn	VBG
zkie-zero-2020	427	2	to	to	IN
zkie-zero-2020	427	3	Link	Link	NNP
zkie-zero-2020	427	4	Entities	Entities	NNP
zkie-zero-2020	427	5	with	with	IN
zkie-zero-2020	427	6	Knowledge	Knowledge	NNP
zkie-zero-2020	427	7	Base	Base	NNP
zkie-zero-2020	427	8	.	.	.
zkie-zero-2020	428	1	In	in	IN
zkie-zero-2020	428	2	Prooceedings	prooceeding	NNS
zkie-zero-2020	428	3	of	of	IN
zkie-zero-2020	428	4	NAACL-	NAACL-	NNP
zkie-zero-2020	428	5	HLT’10	HLT’10	NNP
zkie-zero-2020	428	6	,	,	,
zkie-zero-2020	428	7	pages	page	VBZ
zkie-zero-2020	428	8	483–491	483–491	CD
zkie-zero-2020	428	9	.	.	.
zkie-zero-2020	429	1	A	a	DT
zkie-zero-2020	429	2	Appendices	appendix	NNS
zkie-zero-2020	429	3	A.1	A.1	NNP
zkie-zero-2020	429	4	Dataset	dataset	NN
zkie-zero-2020	429	5	creation	creation	NN
zkie-zero-2020	429	6	The	the	DT
zkie-zero-2020	429	7	following	following	JJ
zkie-zero-2020	429	8	section	section	NN
zkie-zero-2020	429	9	describes	describe	VBZ
zkie-zero-2020	429	10	how	how	WRB
zkie-zero-2020	429	11	we	-PRON-	PRP
zkie-zero-2020	429	12	preprocess	preprocess	VBP
zkie-zero-2020	429	13	the	the	DT
zkie-zero-2020	429	14	raw	raw	JJ
zkie-zero-2020	429	15	texts	text	NNS
zkie-zero-2020	429	16	from	from	IN
zkie-zero-2020	429	17	WWO	WWO	NNP
zkie-zero-2020	429	18	and	and	CC
zkie-zero-2020	429	19	1641	1641	CD
zkie-zero-2020	429	20	.	.	.
zkie-zero-2020	430	1	Example	example	NN
zkie-zero-2020	430	2	texts	text	NNS
zkie-zero-2020	430	3	can	can	MD
zkie-zero-2020	430	4	be	be	VB
zkie-zero-2020	430	5	found	find	VBN
zkie-zero-2020	430	6	in	in	IN
zkie-zero-2020	430	7	Table	table	NN
zkie-zero-2020	430	8	6	6	CD
zkie-zero-2020	430	9	.	.	.
zkie-zero-2020	431	1	The	the	DT
zkie-zero-2020	431	2	respective	respective	JJ
zkie-zero-2020	431	3	code	code	NN
zkie-zero-2020	431	4	and	and	CC
zkie-zero-2020	431	5	datasets	dataset	NNS
zkie-zero-2020	431	6	will	will	MD
zkie-zero-2020	431	7	be	be	VB
zkie-zero-2020	431	8	made	make	VBN
zkie-zero-2020	431	9	available	available	JJ
zkie-zero-2020	431	10	on	on	IN
zkie-zero-2020	431	11	acceptance	acceptance	NN
zkie-zero-2020	431	12	.	.	.
zkie-zero-2020	432	1	A.1.1	A.1.1	NNP
zkie-zero-2020	432	2	Women	woman	NNS
zkie-zero-2020	432	3	Writers	writer	NNS
zkie-zero-2020	432	4	Online	online	RB
zkie-zero-2020	432	5	We	-PRON-	PRP
zkie-zero-2020	432	6	use	use	VBP
zkie-zero-2020	432	7	the	the	DT
zkie-zero-2020	432	8	following	follow	VBG
zkie-zero-2020	432	9	checkout	checkout	NN
zkie-zero-2020	432	10	of	of	IN
zkie-zero-2020	432	11	the	the	DT
zkie-zero-2020	432	12	WWO	WWO	NNP
zkie-zero-2020	432	13	data	datum	NNS
zkie-zero-2020	432	14	,	,	,
zkie-zero-2020	432	15	which	which	WDT
zkie-zero-2020	432	16	was	be	VBD
zkie-zero-2020	432	17	graciously	graciously	RB
zkie-zero-2020	432	18	provided	provide	VBN
zkie-zero-2020	432	19	by	by	IN
zkie-zero-2020	432	20	the	the	DT
zkie-zero-2020	432	21	Women	Women	NNPS
zkie-zero-2020	432	22	Writ-	Writ-	NNP
zkie-zero-2020	432	23	ers	er	VBZ
zkie-zero-2020	432	24	Project6	Project6	NNP
zkie-zero-2020	432	25	.	.	.
zkie-zero-2020	433	1	Revision	revision	NN
zkie-zero-2020	433	2	:	:	:
zkie-zero-2020	433	3	36425	36425	CD
zkie-zero-2020	433	4	Last	last	JJ
zkie-zero-2020	433	5	Changed	Changed	NNP
zkie-zero-2020	433	6	Rev	Rev	NNP
zkie-zero-2020	433	7	:	:	:
zkie-zero-2020	433	8	36341	36341	CD
zkie-zero-2020	433	9	Last	last	JJ
zkie-zero-2020	433	10	Changed	changed	JJ
zkie-zero-2020	433	11	Date	date	NN
zkie-zero-2020	433	12	:	:	:
zkie-zero-2020	433	13	2019	2019	CD
zkie-zero-2020	433	14	-	-	HYPH
zkie-zero-2020	433	15	02	02	CD
zkie-zero-2020	433	16	-	-	HYPH
zkie-zero-2020	433	17	19	19	CD
zkie-zero-2020	433	18	6	6	CD
zkie-zero-2020	433	19	https://www.wwp.northeastern.edu/	https://www.wwp.northeastern.edu/	NN
zkie-zero-2020	433	20	The	the	DT
zkie-zero-2020	433	21	texts	text	NNS
zkie-zero-2020	433	22	itself	-PRON-	PRP
zkie-zero-2020	433	23	are	be	VBP
zkie-zero-2020	433	24	provided	provide	VBN
zkie-zero-2020	433	25	as	as	IN
zkie-zero-2020	433	26	TEI7	TEI7	NNP
zkie-zero-2020	433	27	.	.	.
zkie-zero-2020	434	1	We	-PRON-	PRP
zkie-zero-2020	434	2	use	use	VBP
zkie-zero-2020	434	3	DKPro	DKPro	NNP
zkie-zero-2020	434	4	Core8	Core8	NNS
zkie-zero-2020	434	5	to	to	TO
zkie-zero-2020	434	6	read	read	VB
zkie-zero-2020	434	7	in	in	IN
zkie-zero-2020	434	8	the	the	DT
zkie-zero-2020	434	9	TEI	TEI	NNP
zkie-zero-2020	434	10	,	,	,
zkie-zero-2020	434	11	split	split	VBD
zkie-zero-2020	434	12	the	the	DT
zkie-zero-2020	434	13	raw	raw	JJ
zkie-zero-2020	434	14	text	text	NN
zkie-zero-2020	434	15	into	into	IN
zkie-zero-2020	434	16	sentences	sentence	NNS
zkie-zero-2020	434	17	and	and	CC
zkie-zero-2020	434	18	tokenize	tokenize	VB
zkie-zero-2020	434	19	it	-PRON-	PRP
zkie-zero-2020	434	20	with	with	IN
zkie-zero-2020	434	21	the	the	DT
zkie-zero-2020	434	22	JTokSegmenter	JTokSegmenter	NNP
zkie-zero-2020	434	23	.	.	.
zkie-zero-2020	435	1	When	when	WRB
zkie-zero-2020	435	2	an	an	DT
zkie-zero-2020	435	3	annotation	annotation	NN
zkie-zero-2020	435	4	is	be	VBZ
zkie-zero-2020	435	5	spread	spread	VBN
zkie-zero-2020	435	6	over	over	IN
zkie-zero-2020	435	7	two	two	CD
zkie-zero-2020	435	8	sentences	sentence	NNS
zkie-zero-2020	435	9	,	,	,
zkie-zero-2020	435	10	we	-PRON-	PRP
zkie-zero-2020	435	11	merge	merge	VBP
zkie-zero-2020	435	12	these	these	DT
zkie-zero-2020	435	13	sentences	sentence	NNS
zkie-zero-2020	435	14	.	.	.
zkie-zero-2020	436	1	This	this	DT
zkie-zero-2020	436	2	is	be	VBZ
zkie-zero-2020	436	3	mostly	mostly	RB
zkie-zero-2020	436	4	caused	cause	VBN
zkie-zero-2020	436	5	by	by	IN
zkie-zero-2020	436	6	a	a	DT
zkie-zero-2020	436	7	too	too	RB
zkie-zero-2020	436	8	eager	eager	JJ
zkie-zero-2020	436	9	sentence	sentence	NN
zkie-zero-2020	436	10	splitter	splitter	NNP
zkie-zero-2020	436	11	.	.	.
zkie-zero-2020	437	1	We	-PRON-	PRP
zkie-zero-2020	437	2	covert	covert	VBD
zkie-zero-2020	437	3	the	the	DT
zkie-zero-2020	437	4	personographie	personographie	NN
zkie-zero-2020	437	5	which	which	WDT
zkie-zero-2020	437	6	is	be	VBZ
zkie-zero-2020	437	7	in	in	IN
zkie-zero-2020	437	8	XML	xml	NN
zkie-zero-2020	437	9	to	to	IN
zkie-zero-2020	437	10	RDF	RDF	NNP
zkie-zero-2020	437	11	,	,	,
zkie-zero-2020	437	12	including	include	VBG
zkie-zero-2020	437	13	all	all	DT
zkie-zero-2020	437	14	properties	property	NNS
zkie-zero-2020	437	15	that	that	WDT
zkie-zero-2020	437	16	were	be	VBD
zkie-zero-2020	437	17	encoded	encode	VBN
zkie-zero-2020	437	18	in	in	IN
zkie-zero-2020	437	19	there	there	RB
zkie-zero-2020	437	20	.	.	.
zkie-zero-2020	438	1	A.1.2	A.1.2	NNP
zkie-zero-2020	438	2	1641	1641	CD
zkie-zero-2020	438	3	Depositions	Depositions	NNPS
zkie-zero-2020	438	4	We	-PRON-	PRP
zkie-zero-2020	438	5	use	use	VBP
zkie-zero-2020	438	6	a	a	DT
zkie-zero-2020	438	7	subset	subset	NN
zkie-zero-2020	438	8	of	of	IN
zkie-zero-2020	438	9	the	the	DT
zkie-zero-2020	438	10	1641	1641	CD
zkie-zero-2020	438	11	depositions	deposition	NNS
zkie-zero-2020	438	12	provided	provide	VBN
zkie-zero-2020	438	13	by	by	IN
zkie-zero-2020	438	14	Gary	Gary	NNP
zkie-zero-2020	438	15	Munnelly	munnelly	RB
zkie-zero-2020	438	16	.	.	.
zkie-zero-2020	439	1	The	the	DT
zkie-zero-2020	439	2	raw	raw	JJ
zkie-zero-2020	439	3	data	datum	NNS
zkie-zero-2020	439	4	can	can	MD
zkie-zero-2020	439	5	be	be	VB
zkie-zero-2020	439	6	found	find	VBN
zkie-zero-2020	439	7	on	on	IN
zkie-zero-2020	439	8	Github9	Github9	NNP
zkie-zero-2020	439	9	.	.	.
zkie-zero-2020	440	1	The	the	DT
zkie-zero-2020	440	2	texts	text	NNS
zkie-zero-2020	440	3	itself	-PRON-	PRP
zkie-zero-2020	440	4	are	be	VBP
zkie-zero-2020	440	5	provided	provide	VBN
zkie-zero-2020	440	6	as	as	IN
zkie-zero-2020	440	7	NIF10	NIF10	NNP
zkie-zero-2020	440	8	.	.	.
zkie-zero-2020	441	1	We	-PRON-	PRP
zkie-zero-2020	441	2	use	use	VBP
zkie-zero-2020	441	3	DKPro	DKPro	NNP
zkie-zero-2020	441	4	Core11	Core11	NNP
zkie-zero-2020	441	5	to	to	TO
zkie-zero-2020	441	6	read	read	VB
zkie-zero-2020	441	7	in	in	IN
zkie-zero-2020	441	8	the	the	DT
zkie-zero-2020	441	9	NIF	NIF	NNP
zkie-zero-2020	441	10	,	,	,
zkie-zero-2020	441	11	split	split	VBD
zkie-zero-2020	441	12	the	the	DT
zkie-zero-2020	441	13	raw	raw	JJ
zkie-zero-2020	441	14	text	text	NN
zkie-zero-2020	441	15	into	into	IN
zkie-zero-2020	441	16	sentences	sentence	NNS
zkie-zero-2020	441	17	and	and	CC
zkie-zero-2020	441	18	tokenize	tokenize	VB
zkie-zero-2020	441	19	it	-PRON-	PRP
zkie-zero-2020	441	20	with	with	IN
zkie-zero-2020	441	21	the	the	DT
zkie-zero-2020	441	22	JTokSegmenter	JTokSegmenter	NNP
zkie-zero-2020	441	23	.	.	.
zkie-zero-2020	442	1	When	when	WRB
zkie-zero-2020	442	2	an	an	DT
zkie-zero-2020	442	3	annotation	annotation	NN
zkie-zero-2020	442	4	is	be	VBZ
zkie-zero-2020	442	5	spread	spread	VBN
zkie-zero-2020	442	6	over	over	IN
zkie-zero-2020	442	7	two	two	CD
zkie-zero-2020	442	8	sentences	sentence	NNS
zkie-zero-2020	442	9	,	,	,
zkie-zero-2020	442	10	we	-PRON-	PRP
zkie-zero-2020	442	11	merge	merge	VBP
zkie-zero-2020	442	12	these	these	DT
zkie-zero-2020	442	13	sentences	sentence	NNS
zkie-zero-2020	442	14	.	.	.
zkie-zero-2020	443	1	This	this	DT
zkie-zero-2020	443	2	is	be	VBZ
zkie-zero-2020	443	3	mostly	mostly	RB
zkie-zero-2020	443	4	caused	cause	VBN
zkie-zero-2020	443	5	by	by	IN
zkie-zero-2020	443	6	a	a	DT
zkie-zero-2020	443	7	too	too	RB
zkie-zero-2020	443	8	eager	eager	JJ
zkie-zero-2020	443	9	sentence	sentence	NN
zkie-zero-2020	443	10	splitter	splitter	NNP
zkie-zero-2020	443	11	.	.	.
zkie-zero-2020	444	1	We	-PRON-	PRP
zkie-zero-2020	444	2	use	use	VBP
zkie-zero-2020	444	3	the	the	DT
zkie-zero-2020	444	4	knowledge	knowledge	NN
zkie-zero-2020	444	5	base	base	NN
zkie-zero-2020	444	6	that	that	WDT
zkie-zero-2020	444	7	comes	come	VBZ
zkie-zero-2020	444	8	with	with	IN
zkie-zero-2020	444	9	the	the	DT
zkie-zero-2020	444	10	NIF	NIF	NNP
zkie-zero-2020	444	11	and	and	CC
zkie-zero-2020	444	12	create	create	VB
zkie-zero-2020	444	13	entities	entity	NNS
zkie-zero-2020	444	14	for	for	IN
zkie-zero-2020	444	15	all	all	DT
zkie-zero-2020	444	16	mentions	mention	NNS
zkie-zero-2020	444	17	that	that	WDT
zkie-zero-2020	444	18	were	be	VBD
zkie-zero-2020	444	19	NIL	NIL	NNP
zkie-zero-2020	444	20	.	.	.
zkie-zero-2020	445	1	We	-PRON-	PRP
zkie-zero-2020	445	2	carefully	carefully	RB
zkie-zero-2020	445	3	deduplicate	deduplicate	VBP
zkie-zero-2020	445	4	entities	entity	NNS
zkie-zero-2020	445	5	,	,	,
zkie-zero-2020	445	6	e.g.	e.g.	RB
zkie-zero-2020	446	1	Luke	Luke	NNP
zkie-zero-2020	446	2	Toole	Toole	NNP
zkie-zero-2020	446	3	and	and	CC
zkie-zero-2020	446	4	Colonel	Colonel	NNP
zkie-zero-2020	446	5	Toole	Toole	NNP
zkie-zero-2020	446	6	are	be	VBP
zkie-zero-2020	446	7	mapped	map	VBN
zkie-zero-2020	446	8	to	to	IN
zkie-zero-2020	446	9	the	the	DT
zkie-zero-2020	446	10	same	same	JJ
zkie-zero-2020	446	11	entity	entity	NN
zkie-zero-2020	446	12	.	.	.
zkie-zero-2020	447	1	In	in	IN
zkie-zero-2020	447	2	order	order	NN
zkie-zero-2020	447	3	to	to	TO
zkie-zero-2020	447	4	increase	increase	VB
zkie-zero-2020	447	5	the	the	DT
zkie-zero-2020	447	6	difficulty	difficulty	NN
zkie-zero-2020	447	7	of	of	IN
zkie-zero-2020	447	8	this	this	DT
zkie-zero-2020	447	9	dataset	dataset	NN
zkie-zero-2020	447	10	,	,	,
zkie-zero-2020	447	11	we	-PRON-	PRP
zkie-zero-2020	447	12	add	add	VBP
zkie-zero-2020	447	13	additional	additional	JJ
zkie-zero-2020	447	14	entities	entity	NNS
zkie-zero-2020	447	15	from	from	IN
zkie-zero-2020	447	16	DB-	DB-	NNP
zkie-zero-2020	447	17	Pedia	Pedia	NNP
zkie-zero-2020	447	18	:	:	:
zkie-zero-2020	447	19	all	all	DT
zkie-zero-2020	447	20	Irish	irish	JJ
zkie-zero-2020	447	21	people	people	NNS
zkie-zero-2020	447	22	,	,	,
zkie-zero-2020	447	23	Irish	irish	JJ
zkie-zero-2020	447	24	cities	city	NNS
zkie-zero-2020	447	25	and	and	CC
zkie-zero-2020	447	26	buildings	building	NNS
zkie-zero-2020	447	27	in	in	IN
zkie-zero-2020	447	28	Ireland	Ireland	NNP
zkie-zero-2020	447	29	;	;	:
zkie-zero-2020	447	30	all	all	DT
zkie-zero-2020	447	31	popes	pope	NNS
zkie-zero-2020	447	32	;	;	:
zkie-zero-2020	447	33	royalities	royalitie	NNS
zkie-zero-2020	447	34	born	bear	VBN
zkie-zero-2020	447	35	between	between	IN
zkie-zero-2020	447	36	1550	1550	CD
zkie-zero-2020	447	37	and	and	CC
zkie-zero-2020	447	38	1650	1650	CD
zkie-zero-2020	447	39	.	.	.
zkie-zero-2020	448	1	For	for	IN
zkie-zero-2020	448	2	that	that	DT
zkie-zero-2020	448	3	,	,	,
zkie-zero-2020	448	4	we	-PRON-	PRP
zkie-zero-2020	448	5	execute	execute	VBP
zkie-zero-2020	448	6	SPARQL	sparql	JJ
zkie-zero-2020	448	7	queries	query	NNS
zkie-zero-2020	448	8	against	against	IN
zkie-zero-2020	448	9	DBPedia	DBPedia	NNP
zkie-zero-2020	448	10	for	for	IN
zkie-zero-2020	448	11	instances	instance	NNS
zkie-zero-2020	448	12	of	of	IN
zkie-zero-2020	448	13	dbc	dbc	NN
zkie-zero-2020	448	14	:	:	:
zkie-zero-2020	448	15	Popes	Popes	NNP
zkie-zero-2020	448	16	,	,	,
zkie-zero-2020	448	17	dbc	dbc	NN
zkie-zero-2020	448	18	:	:	:
zkie-zero-2020	448	19	Royality	royality	NN
zkie-zero-2020	448	20	,	,	,
zkie-zero-2020	448	21	dbc:17th	dbc:17th	JJ
zkie-zero-2020	448	22	-	-	HYPH
zkie-zero-2020	448	23	century	century	NN
zkie-zero-2020	448	24	Irish	irish	JJ
zkie-zero-2020	448	25	people	people	NNS
zkie-zero-2020	448	26	and	and	CC
zkie-zero-2020	448	27	keep	keep	VB
zkie-zero-2020	448	28	entries	entry	NNS
zkie-zero-2020	448	29	with	with	IN
zkie-zero-2020	448	30	a	a	DT
zkie-zero-2020	448	31	birth	birth	NN
zkie-zero-2020	448	32	date	date	NN
zkie-zero-2020	448	33	before	before	IN
zkie-zero-2020	448	34	1650	1650	CD
zkie-zero-2020	448	35	and	and	CC
zkie-zero-2020	448	36	a	a	DT
zkie-zero-2020	448	37	death	death	NN
zkie-zero-2020	448	38	date	date	NN
zkie-zero-2020	448	39	between	between	IN
zkie-zero-2020	448	40	1600	1600	CD
zkie-zero-2020	448	41	and	and	CC
zkie-zero-2020	448	42	1700	1700	CD
zkie-zero-2020	448	43	.	.	.
zkie-zero-2020	449	1	For	for	IN
zkie-zero-2020	449	2	the	the	DT
zkie-zero-2020	449	3	places	place	NNS
zkie-zero-2020	449	4	,	,	,
zkie-zero-2020	449	5	we	-PRON-	PRP
zkie-zero-2020	449	6	search	search	VBP
zkie-zero-2020	449	7	for	for	IN
zkie-zero-2020	449	8	dbo	dbo	NNP
zkie-zero-2020	449	9	:	:	:
zkie-zero-2020	449	10	Castle	Castle	NNP
zkie-zero-2020	449	11	,	,	,
zkie-zero-2020	449	12	dbo	dbo	NN
zkie-zero-2020	449	13	:	:	:
zkie-zero-2020	449	14	HistoricPlace	HistoricPlace	NNP
zkie-zero-2020	449	15	,	,	,
zkie-zero-2020	449	16	dbo	dbo	NNP
zkie-zero-2020	449	17	:	:	:
zkie-zero-2020	449	18	Building	building	NN
zkie-zero-2020	449	19	,	,	,
zkie-zero-2020	449	20	dbc:17th	dbc:17th	JJ
zkie-zero-2020	449	21	-	-	HYPH
zkie-zero-2020	449	22	century	century	NN
zkie-zero-2020	449	23	Irish	irish	JJ
zkie-zero-2020	449	24	people	people	NNS
zkie-zero-2020	449	25	that	that	WDT
zkie-zero-2020	449	26	are	be	VBP
zkie-zero-2020	449	27	located	locate	VBN
zkie-zero-2020	449	28	in	in	IN
zkie-zero-2020	449	29	Ireland	Ireland	NNP
zkie-zero-2020	449	30	.	.	.
zkie-zero-2020	450	1	The	the	DT
zkie-zero-2020	450	2	follwing	follwe	VBG
zkie-zero-2020	450	3	table	table	NN
zkie-zero-2020	450	4	shows	show	VBZ
zkie-zero-2020	450	5	how	how	WRB
zkie-zero-2020	450	6	many	many	JJ
zkie-zero-2020	450	7	entities	entity	NNS
zkie-zero-2020	450	8	were	be	VBD
zkie-zero-2020	450	9	in	in	IN
zkie-zero-2020	450	10	the	the	DT
zkie-zero-2020	450	11	original	original	JJ
zkie-zero-2020	450	12	KB	KB	NNP
zkie-zero-2020	450	13	and	and	CC
zkie-zero-2020	450	14	how	how	WRB
zkie-zero-2020	450	15	many	many	JJ
zkie-zero-2020	450	16	were	be	VBD
zkie-zero-2020	450	17	added	add	VBN
zkie-zero-2020	450	18	:	:	:
zkie-zero-2020	450	19	Persons	person	NNS
zkie-zero-2020	450	20	in	in	IN
zkie-zero-2020	450	21	gold	gold	JJ
zkie-zero-2020	450	22	data	datum	NNS
zkie-zero-2020	450	23	130	130	CD
zkie-zero-2020	450	24	Places	Places	NNPS
zkie-zero-2020	450	25	in	in	IN
zkie-zero-2020	450	26	gold	gold	NN
zkie-zero-2020	450	27	data	datum	NNS
zkie-zero-2020	450	28	99	99	CD
zkie-zero-2020	450	29	Persons	Persons	NNPS
zkie-zero-2020	450	30	added	add	VBD
zkie-zero-2020	450	31	from	from	IN
zkie-zero-2020	450	32	DBPedia	DBPedia	NNP
zkie-zero-2020	450	33	1253	1253	CD
zkie-zero-2020	450	34	Places	Places	NNPS
zkie-zero-2020	450	35	added	add	VBN
zkie-zero-2020	450	36	from	from	IN
zkie-zero-2020	450	37	DBPedia	DBPedia	NNP
zkie-zero-2020	450	38	2020	2020	CD
zkie-zero-2020	450	39	7	7	CD
zkie-zero-2020	450	40	https://tei-c.org/	https://tei-c.org/	NNP
zkie-zero-2020	450	41	8	8	CD
zkie-zero-2020	450	42	https://dkpro.github.io/dkpro-core/	https://dkpro.github.io/dkpro-core/	NNP
zkie-zero-2020	450	43	9	9	CD
zkie-zero-2020	450	44	https://github.com/munnellg/	https://github.com/munnellg/	CD
zkie-zero-2020	450	45	1641DepositionsCorpus	1641depositionscorpus	CD
zkie-zero-2020	450	46	10	10	CD
zkie-zero-2020	450	47	https://persistence.uni-leipzig.org/	https://persistence.uni-leipzig.org/	NN
zkie-zero-2020	450	48	nlp2rdf/	nlp2rdf/	VBZ
zkie-zero-2020	450	49	11	11	CD
zkie-zero-2020	450	50	https://dkpro.github.io/dkpro-core/	https://dkpro.github.io/dkpro-core/	NN
zkie-zero-2020	450	51	https://doi.org/10.1007/s10032-002-0082-8	https://doi.org/10.1007/s10032-002-0082-8	CD
zkie-zero-2020	450	52	https://doi.org/10.1007/s10032-002-0082-8	https://doi.org/10.1007/s10032-002-0082-8	CD
zkie-zero-2020	450	53	https://doi.org/10.1109/tkde.2014.2327028	https://doi.org/10.1109/tkde.2014.2327028	NNP
zkie-zero-2020	450	54	https://doi.org/10.1109/tkde.2014.2327028	https://doi.org/10.1109/tkde.2014.2327028	NNP
zkie-zero-2020	450	55	https://doi.org/10.1109/tkde.2014.2327028	https://doi.org/10.1109/tkde.2014.2327028	NNP
zkie-zero-2020	450	56	https://doi.org/10.1007/978-3-319-11964-9_29	https://doi.org/10.1007/978-3-319-11964-9_29	NNP
zkie-zero-2020	450	57	https://doi.org/10.1007/978-3-319-11964-9_29	https://doi.org/10.1007/978-3-319-11964-9_29	NNP
zkie-zero-2020	450	58	https://transacl.org/ojs/index.php/tacl/article/view/1711	https://transacl.org/ojs/index.php/tacl/article/view/1711	,
zkie-zero-2020	450	59	https://transacl.org/ojs/index.php/tacl/article/view/1711	https://transacl.org/ojs/index.php/tacl/article/view/1711	:
zkie-zero-2020	450	60	https://transacl.org/ojs/index.php/tacl/article/view/1711	https://transacl.org/ojs/index.php/tacl/article/view/1711	``
zkie-zero-2020	450	61	http://arxiv.org/abs/1506.00301	http://arxiv.org/abs/1506.00301	NNP
zkie-zero-2020	450	62	http://arxiv.org/abs/1506.00301	http://arxiv.org/abs/1506.00301	NNP
zkie-zero-2020	450	63	https://doi.org/10.3115/v1/p14-5016	https://doi.org/10.3115/v1/p14-5016	RB
zkie-zero-2020	450	64	https://doi.org/10.3115/v1/p14-5016	https://doi.org/10.3115/v1/p14-5016	RB
zkie-zero-2020	450	65	https://doi.org/10.3115/v1/p14-5016	https://doi.org/10.3115/v1/p14-5016	JJ
zkie-zero-2020	450	66	https://www.aclweb.org/anthology/N10-1072.pdf	https://www.aclweb.org/anthology/n10-1072.pdf	NN
zkie-zero-2020	450	67	https://www.aclweb.org/anthology/N10-1072.pdf	https://www.aclweb.org/anthology/N10-1072.pdf	NNP
zkie-zero-2020	450	68	https://www.wwp.northeastern.edu/	https://www.wwp.northeastern.edu/	NNP
zkie-zero-2020	450	69	https://tei-c.org/	https://tei-c.org/	NNP
zkie-zero-2020	450	70	https://dkpro.github.io/dkpro-core/	https://dkpro.github.io/dkpro-core/	NNP
zkie-zero-2020	450	71	https://github.com/munnellg/1641DepositionsCorpus	https://github.com/munnellg/1641DepositionsCorpus	:
zkie-zero-2020	450	72	https://github.com/munnellg/1641DepositionsCorpus	https://github.com/munnellg/1641depositionscorpus	JJ
zkie-zero-2020	450	73	https://persistence.uni-leipzig.org/nlp2rdf/	https://persistence.uni-leipzig.org/nlp2rdf/	NN
zkie-zero-2020	450	74	https://persistence.uni-leipzig.org/nlp2rdf/	https://persistence.uni-leipzig.org/nlp2rdf/	NNS
zkie-zero-2020	450	75	https://dkpro.github.io/dkpro-core/	https://dkpro.github.io/dkpro-core/	NNP
zkie-zero-2020	450	76	6993	6993	CD
zkie-zero-2020	450	77	WWO	wwo	NN
zkie-zero-2020	450	78	The	the	DT
zkie-zero-2020	450	79	following	follow	VBG
zkie-zero-2020	450	80	Lines	line	NNS
zkie-zero-2020	450	81	occasion’d	occasion’d	VBN
zkie-zero-2020	450	82	by	by	IN
zkie-zero-2020	450	83	the	the	DT
zkie-zero-2020	450	84	Marriage	Marriage	NNP
zkie-zero-2020	450	85	of	of	IN
zkie-zero-2020	450	86	Edward	Edward	NNP
zkie-zero-2020	450	87	Herbert	Herbert	NNP
zkie-zero-2020	450	88	Esquire	Esquire	NNP
zkie-zero-2020	450	89	,	,	,
zkie-zero-2020	450	90	and	and	CC
zkie-zero-2020	450	91	Mrs.	Mrs.	NNP
zkie-zero-2020	450	92	Eliza-	Eliza-	NNP
zkie-zero-2020	450	93	beth	beth	NN
zkie-zero-2020	450	94	Herbert	Herbert	NNP
zkie-zero-2020	450	95	.	.	.
zkie-zero-2020	451	1	Cupid	Cupid	NNP
zkie-zero-2020	451	2	one	one	CD
zkie-zero-2020	451	3	day	day	NN
zkie-zero-2020	451	4	ask’d	ask’d	NNP
zkie-zero-2020	451	5	his	-PRON-	PRP$
zkie-zero-2020	451	6	Mother	mother	NN
zkie-zero-2020	451	7	,	,	,
zkie-zero-2020	451	8	When	when	WRB
zkie-zero-2020	451	9	she	-PRON-	PRP
zkie-zero-2020	451	10	meant	mean	VBD
zkie-zero-2020	451	11	that	that	IN
zkie-zero-2020	451	12	he	-PRON-	PRP
zkie-zero-2020	451	13	shou’d	shou’d	VB
zkie-zero-2020	451	14	We	-PRON-	PRP
zkie-zero-2020	451	15	d	d	NN
zkie-zero-2020	451	16	?	?	.
zkie-zero-2020	452	1	You	-PRON-	PRP
zkie-zero-2020	452	2	’re	be	VBZ
zkie-zero-2020	452	3	too	too	RB
zkie-zero-2020	452	4	Young	young	JJ
zkie-zero-2020	452	5	,	,	,
zkie-zero-2020	452	6	my	-PRON-	PRP$
zkie-zero-2020	452	7	Boy	boy	UH
zkie-zero-2020	452	8	,	,	,
zkie-zero-2020	452	9	she	-PRON-	PRP
zkie-zero-2020	452	10	said	say	VBD
zkie-zero-2020	452	11	:	:	:
zkie-zero-2020	452	12	Nor	nor	CC
zkie-zero-2020	452	13	has	have	VBZ
zkie-zero-2020	452	14	Nature	nature	NN
zkie-zero-2020	452	15	made	make	VBN
zkie-zero-2020	452	16	another	another	DT
zkie-zero-2020	452	17	Fit	Fit	NNP
zkie-zero-2020	452	18	to	to	TO
zkie-zero-2020	452	19	match	match	VB
zkie-zero-2020	452	20	with	with	IN
zkie-zero-2020	452	21	Cupid	Cupid	NNP
zkie-zero-2020	452	22	’s	’s	,
zkie-zero-2020	452	23	Bed	bed	NN
zkie-zero-2020	452	24	.	.	.
zkie-zero-2020	453	1	Finch	Finch	NNP
zkie-zero-2020	453	2	,	,	,
zkie-zero-2020	453	3	Anne	Anne	NNP
zkie-zero-2020	453	4	:	:	:
zkie-zero-2020	453	5	Miscellany	Miscellany	NNP
zkie-zero-2020	453	6	poems	poem	NNS
zkie-zero-2020	453	7	,	,	,
zkie-zero-2020	453	8	on	on	IN
zkie-zero-2020	453	9	several	several	JJ
zkie-zero-2020	453	10	occasions	occasion	NNS
zkie-zero-2020	453	11	,	,	,
zkie-zero-2020	453	12	1713	1713	CD
zkie-zero-2020	453	13	Joseph	Joseph	NNP
zkie-zero-2020	453	14	Joice	Joice	NNP
zkie-zero-2020	453	15	of	of	IN
zkie-zero-2020	453	16	Kisnebrasney	Kisnebrasney	NNP
zkie-zero-2020	453	17	in	in	IN
zkie-zero-2020	453	18	the	the	DT
zkie-zero-2020	453	19	kings	king	NNS
zkie-zero-2020	453	20	County	County	NNP
zkie-zero-2020	453	21	gentleman	gentleman	NN
zkie-zero-2020	453	22	sworne	sworne	NN
zkie-zero-2020	453	23	and	and	CC
zkie-zero-2020	453	24	examined	examine	VBD
zkie-zero-2020	453	25	de-	de-	CC
zkie-zero-2020	453	26	poseth	poseth	NN
zkie-zero-2020	453	27	and	and	CC
zkie-zero-2020	453	28	saith	saith	JJ
zkie-zero-2020	453	29	That	that	IN
zkie-zero-2020	453	30	after	after	IN
zkie-zero-2020	453	31	the	the	DT
zkie-zero-2020	453	32	Rebellion	rebellion	NN
zkie-zero-2020	453	33	was	be	VBD
zkie-zero-2020	453	34	begun	begin	VBN
zkie-zero-2020	453	35	in	in	IN
zkie-zero-2020	453	36	the	the	DT
zkie-zero-2020	453	37	County	County	NNP
zkie-zero-2020	453	38	aforesaid	aforesaid	VBD
zkie-zero-2020	453	39	vizt	vizt	NN
zkie-zero-2020	453	40	about	about	IN
zkie-zero-2020	453	41	the	the	DT
zkie-zero-2020	453	42	xxth	xxth	NN
zkie-zero-2020	453	43	of	of	IN
zkie-zero-2020	453	44	November	November	NNP
zkie-zero-2020	453	45	1641	1641	CD
zkie-zero-2020	453	46	This	this	DT
zkie-zero-2020	453	47	deponent	deponent	NN
zkie-zero-2020	453	48	for	for	IN
zkie-zero-2020	453	49	saffty	saffty	NN
zkie-zero-2020	453	50	fled	flee	VBD
zkie-zero-2020	453	51	to	to	IN
zkie-zero-2020	453	52	the	the	DT
zkie-zero-2020	453	53	Castle	Castle	NNP
zkie-zero-2020	453	54	of	of	IN
zkie-zero-2020	453	55	knocknamease	knocknamease	NN
zkie-zero-2020	453	56	in	in	IN
zkie-zero-2020	453	57	the	the	DT
zkie-zero-2020	453	58	same	same	JJ
zkie-zero-2020	453	59	County	County	NNP
zkie-zero-2020	453	60	Deposition	Deposition	NNP
zkie-zero-2020	453	61	of	of	IN
zkie-zero-2020	453	62	Joseph	Joseph	NNP
zkie-zero-2020	453	63	Joice	Joice	NNP
zkie-zero-2020	453	64	,	,	,
zkie-zero-2020	453	65	164312	164312	CD
zkie-zero-2020	453	66	Table	table	NN
zkie-zero-2020	453	67	6	6	CD
zkie-zero-2020	453	68	:	:	:
zkie-zero-2020	453	69	Example	example	NN
zkie-zero-2020	453	70	sentences	sentence	NNS
zkie-zero-2020	453	71	from	from	IN
zkie-zero-2020	453	72	these	these	DT
zkie-zero-2020	453	73	corpora	corpora	NN
zkie-zero-2020	453	74	.	.	.
zkie-zero-2020	454	1	Linked	link	VBN
zkie-zero-2020	454	2	Named	name	VBN
zkie-zero-2020	454	3	entities	entity	NNS
zkie-zero-2020	454	4	are	be	VBP
zkie-zero-2020	454	5	highlighted	highlight	VBN
zkie-zero-2020	454	6	in	in	IN
zkie-zero-2020	454	7	yellow	yellow	NN
zkie-zero-2020	454	8	.	.	.
zkie-zero-2020	455	1	A.2	a.2	NN
zkie-zero-2020	455	2	Experiments	experiment	NNS
zkie-zero-2020	455	3	A.2.1	A.2.1	NNP
zkie-zero-2020	455	4	Full	full	JJ
zkie-zero-2020	455	5	text	text	NN
zkie-zero-2020	455	6	search	search	NN
zkie-zero-2020	455	7	For	for	IN
zkie-zero-2020	455	8	AIDA	AIDA	NNP
zkie-zero-2020	455	9	and	and	CC
zkie-zero-2020	455	10	Wikidata	Wikidata	NNP
zkie-zero-2020	455	11	,	,	,
zkie-zero-2020	455	12	we	-PRON-	PRP
zkie-zero-2020	455	13	use	use	VBP
zkie-zero-2020	455	14	the	the	DT
zkie-zero-2020	455	15	official	official	JJ
zkie-zero-2020	455	16	SPARQL	sparql	NN
zkie-zero-2020	455	17	endpoint	endpoint	NN
zkie-zero-2020	455	18	and	and	CC
zkie-zero-2020	455	19	the	the	DT
zkie-zero-2020	455	20	Mediawiki	Mediawiki	NNP
zkie-zero-2020	455	21	API	API	NNP
zkie-zero-2020	455	22	Query	Query	NNP
zkie-zero-2020	455	23	Service13	Service13	NNP
zkie-zero-2020	455	24	.	.	.
zkie-zero-2020	456	1	It	-PRON-	PRP
zkie-zero-2020	456	2	does	do	VBZ
zkie-zero-2020	456	3	not	not	RB
zkie-zero-2020	456	4	support	support	VB
zkie-zero-2020	456	5	fuzzy	fuzzy	JJ
zkie-zero-2020	456	6	search	search	NN
zkie-zero-2020	456	7	.	.	.
zkie-zero-2020	457	1	For	for	IN
zkie-zero-2020	457	2	WWO	WWO	NNP
zkie-zero-2020	457	3	and	and	CC
zkie-zero-2020	457	4	1641	1641	CD
zkie-zero-2020	457	5	,	,	,
zkie-zero-2020	457	6	we	-PRON-	PRP
zkie-zero-2020	457	7	host	host	VBP
zkie-zero-2020	457	8	the	the	DT
zkie-zero-2020	457	9	created	create	VBN
zkie-zero-2020	457	10	RDF	RDF	NNP
zkie-zero-2020	457	11	in	in	IN
zkie-zero-2020	457	12	a	a	DT
zkie-zero-2020	457	13	Fuseki14	Fuseki14	NNP
zkie-zero-2020	457	14	instance	instance	NN
zkie-zero-2020	457	15	and	and	CC
zkie-zero-2020	457	16	use	use	VB
zkie-zero-2020	457	17	the	the	DT
zkie-zero-2020	457	18	builtin	builtin	NNP
zkie-zero-2020	457	19	func-	func-	NNP
zkie-zero-2020	457	20	tionality	tionality	NN
zkie-zero-2020	457	21	to	to	TO
zkie-zero-2020	457	22	index	index	NN
zkie-zero-2020	457	23	via	via	IN
zkie-zero-2020	457	24	Lucene	Lucene	NNP
zkie-zero-2020	457	25	.	.	.
zkie-zero-2020	458	1	A.2.2	A.2.2	NNP
zkie-zero-2020	458	2	Timing	Timing	NNP
zkie-zero-2020	458	3	Timing	Timing	NNP
zkie-zero-2020	458	4	was	be	VBD
zkie-zero-2020	458	5	performed	perform	VBN
zkie-zero-2020	458	6	on	on	IN
zkie-zero-2020	458	7	a	a	DT
zkie-zero-2020	458	8	Desktop	Desktop	NNP
zkie-zero-2020	458	9	PC	pc	NN
zkie-zero-2020	458	10	with	with	IN
zkie-zero-2020	458	11	Ryzen	Ryzen	NNP
zkie-zero-2020	458	12	3600	3600	CD
zkie-zero-2020	458	13	and	and	CC
zkie-zero-2020	458	14	a	a	DT
zkie-zero-2020	458	15	GeForce	GeForce	NNP
zkie-zero-2020	458	16	RTX	RTX	NNP
zkie-zero-2020	458	17	2060	2060	CD
zkie-zero-2020	458	18	.	.	.
zkie-zero-2020	459	1	13	13	CD
zkie-zero-2020	459	2	https://www.mediawiki.org/wiki/	https://www.mediawiki.org/wiki/	NNP
zkie-zero-2020	459	3	Wikidata_Query_Service	Wikidata_Query_Service	NNP
zkie-zero-2020	459	4	/	/	SYM
zkie-zero-2020	459	5	User_Manual	user_manual	CD
zkie-zero-2020	459	6	/	/	SYM
zkie-zero-2020	459	7	MWAPI	MWAPI	NNP
zkie-zero-2020	459	8	14	14	CD
zkie-zero-2020	459	9	https://jena.apache.org/	https://jena.apache.org/	NN
zkie-zero-2020	459	10	documentation	documentation	NN
zkie-zero-2020	459	11	/	/	SYM
zkie-zero-2020	459	12	fuseki2/	fuseki2/	ADD
zkie-zero-2020	459	13	https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual/MWAPI	https://www.mediawiki.org/wiki/wikidata_query_service/user_manual/mwapi	ADD
zkie-zero-2020	459	14	https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual/MWAPI	https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual/MWAPI	NNP
zkie-zero-2020	459	15	https://jena.apache.org/documentation/fuseki2/	https://jena.apache.org/documentation/fuseki2/	NNP
zkie-zero-2020	459	16	https://jena.apache.org/documentation/fuseki2/	https://jena.apache.org/documentation/fuseki2/	NNP