id	sid	tid	token	lemma	pos
fcis-6912	1	1	frontiers	frontier	NOUN
fcis-6912	1	2	in	in	ADP
fcis-6912	1	3	computing	computing	NOUN
fcis-6912	1	4	and	and	CCONJ
fcis-6912	1	5	intelligent	intelligent	ADJ
fcis-6912	1	6	systems	system	NOUN
fcis-6912	1	7	issn	issn	VERB
fcis-6912	1	8	:	:	PUNCT
fcis-6912	1	9	2832	2832	NUM
fcis-6912	1	10	-	-	SYM
fcis-6912	1	11	6024	6024	NUM
fcis-6912	1	12	|	|	NOUN
fcis-6912	1	13	vol	vol	NOUN
fcis-6912	1	14	.	.	PROPN
fcis-6912	2	1	3	3	NUM
fcis-6912	2	2	,	,	PUNCT
fcis-6912	2	3	no	no	INTJ
fcis-6912	2	4	.	.	NOUN
fcis-6912	2	5	2	2	NUM
fcis-6912	2	6	,	,	PUNCT
fcis-6912	2	7	2023	2023	NUM
fcis-6912	2	8	21	21	NUM
fcis-6912	2	9	chinese	chinese	PROPN
fcis-6912	2	10	named	name	VERB
fcis-6912	2	11	entity	entity	NOUN
fcis-6912	2	12	recognition	recognition	NOUN
fcis-6912	2	13	based	base	VERB
fcis-6912	2	14	on	on	ADP
fcis-6912	2	15	ernie	ernie	PROPN
fcis-6912	2	16	xing	xing	PROPN
fcis-6912	2	17	qi	qi	PROPN
fcis-6912	2	18	school	school	NOUN
fcis-6912	2	19	of	of	ADP
fcis-6912	2	20	electrical	electrical	ADJ
fcis-6912	2	21	engineering	engineering	NOUN
fcis-6912	2	22	,	,	PUNCT
fcis-6912	2	23	southwest	southwest	PROPN
fcis-6912	2	24	minzu	minzu	PROPN
fcis-6912	2	25	university	university	PROPN
fcis-6912	2	26	,	,	PUNCT
fcis-6912	2	27	chengdu	chengdu	PROPN
fcis-6912	2	28	,	,	PUNCT
fcis-6912	2	29	china	china	PROPN
fcis-6912	2	30	abstract	abstract	PROPN
fcis-6912	2	31	:	:	PUNCT
fcis-6912	2	32	the	the	DET
fcis-6912	2	33	traditional	traditional	ADJ
fcis-6912	2	34	named	name	VERB
fcis-6912	2	35	entity	entity	NOUN
fcis-6912	2	36	recognition	recognition	NOUN
fcis-6912	2	37	model	model	NOUN
fcis-6912	2	38	based	base	VERB
fcis-6912	2	39	on	on	ADP
fcis-6912	2	40	neural	neural	ADJ
fcis-6912	2	41	network	network	NOUN
fcis-6912	2	42	uses	use	VERB
fcis-6912	2	43	static	static	ADJ
fcis-6912	2	44	word	word	NOUN
fcis-6912	2	45	vector	vector	NOUN
fcis-6912	2	46	,	,	PUNCT
fcis-6912	2	47	which	which	PRON
fcis-6912	2	48	ca	can	AUX
fcis-6912	2	49	n’t	not	PART
fcis-6912	2	50	represent	represent	VERB
fcis-6912	2	51	the	the	DET
fcis-6912	2	52	ambiguity	ambiguity	NOUN
fcis-6912	2	53	of	of	ADP
fcis-6912	2	54	the	the	DET
fcis-6912	2	55	word	word	NOUN
fcis-6912	2	56	in	in	ADP
fcis-6912	2	57	the	the	DET
fcis-6912	2	58	context	context	NOUN
fcis-6912	2	59	.	.	PUNCT
fcis-6912	3	1	the	the	DET
fcis-6912	3	2	ernie	ernie	PROPN
fcis-6912	3	3	-	-	PUNCT
fcis-6912	3	4	bilstm	bilstm	NOUN
fcis-6912	3	5	-	-	PUNCT
fcis-6912	3	6	crf	crf	NOUN
fcis-6912	3	7	model	model	NOUN
fcis-6912	3	8	is	be	AUX
fcis-6912	3	9	proposed	propose	VERB
fcis-6912	3	10	.	.	PUNCT
fcis-6912	4	1	the	the	DET
fcis-6912	4	2	ernie	ernie	PROPN
fcis-6912	4	3	pre	pre	ADJ
fcis-6912	4	4	-	-	ADJ
fcis-6912	4	5	training	training	ADJ
fcis-6912	4	6	model	model	NOUN
fcis-6912	4	7	can	can	AUX
fcis-6912	4	8	output	output	VERB
fcis-6912	4	9	different	different	ADJ
fcis-6912	4	10	word	word	NOUN
fcis-6912	4	11	vectors	vector	NOUN
fcis-6912	4	12	for	for	ADP
fcis-6912	4	13	different	different	ADJ
fcis-6912	4	14	contexts	context	NOUN
fcis-6912	4	15	by	by	ADP
fcis-6912	4	16	using	use	VERB
fcis-6912	4	17	multiple	multiple	ADJ
fcis-6912	4	18	layers	layer	NOUN
fcis-6912	4	19	of	of	ADP
fcis-6912	4	20	transformer	transformer	NOUN
fcis-6912	4	21	,	,	PUNCT
fcis-6912	4	22	obtaining	obtain	VERB
fcis-6912	4	23	dynamic	dynamic	ADJ
fcis-6912	4	24	word	word	NOUN
fcis-6912	4	25	vectors	vector	NOUN
fcis-6912	4	26	that	that	PRON
fcis-6912	4	27	contain	contain	VERB
fcis-6912	4	28	overall	overall	ADJ
fcis-6912	4	29	sequence	sequence	NOUN
fcis-6912	4	30	information	information	NOUN
fcis-6912	4	31	.	.	PUNCT
fcis-6912	5	1	secondly	secondly	ADV
fcis-6912	5	2	,	,	PUNCT
fcis-6912	5	3	the	the	DET
fcis-6912	5	4	word	word	NOUN
fcis-6912	5	5	vectors	vector	NOUN
fcis-6912	5	6	are	be	AUX
fcis-6912	5	7	input	input	VERB
fcis-6912	5	8	into	into	ADP
fcis-6912	5	9	the	the	DET
fcis-6912	5	10	bilstm	bilstm	NOUN
fcis-6912	5	11	layer	layer	NOUN
fcis-6912	5	12	,	,	PUNCT
fcis-6912	5	13	which	which	PRON
fcis-6912	5	14	can	can	AUX
fcis-6912	5	15	obtain	obtain	VERB
fcis-6912	5	16	sentence	sentence	NOUN
fcis-6912	5	17	context	context	NOUN
fcis-6912	5	18	information	information	NOUN
fcis-6912	5	19	through	through	ADP
fcis-6912	5	20	forward	forward	ADV
fcis-6912	5	21	and	and	CCONJ
fcis-6912	5	22	backward	backward	ADJ
fcis-6912	5	23	lstm	lstm	NOUN
fcis-6912	5	24	and	and	CCONJ
fcis-6912	5	25	obtain	obtain	VERB
fcis-6912	5	26	more	more	ADJ
fcis-6912	5	27	sentence	sentence	NOUN
fcis-6912	5	28	features	feature	NOUN
fcis-6912	5	29	,	,	PUNCT
fcis-6912	5	30	thereby	thereby	ADV
fcis-6912	5	31	improving	improve	VERB
fcis-6912	5	32	the	the	DET
fcis-6912	5	33	model	model	NOUN
fcis-6912	5	34	's	's	PART
fcis-6912	5	35	effectiveness	effectiveness	NOUN
fcis-6912	5	36	.	.	PUNCT
fcis-6912	6	1	finally	finally	ADV
fcis-6912	6	2	,	,	PUNCT
fcis-6912	6	3	the	the	DET
fcis-6912	6	4	sequence	sequence	NOUN
fcis-6912	6	5	is	be	AUX
fcis-6912	6	6	labeled	label	VERB
fcis-6912	6	7	through	through	ADP
fcis-6912	6	8	the	the	DET
fcis-6912	6	9	crf	crf	NOUN
fcis-6912	6	10	layer	layer	NOUN
fcis-6912	6	11	to	to	PART
fcis-6912	6	12	obtain	obtain	VERB
fcis-6912	6	13	the	the	DET
fcis-6912	6	14	globally	globally	ADV
fcis-6912	6	15	optimal	optimal	ADJ
fcis-6912	6	16	labeling	labeling	NOUN
fcis-6912	6	17	information	information	NOUN
fcis-6912	6	18	and	and	CCONJ
fcis-6912	6	19	complete	complete	VERB
fcis-6912	6	20	the	the	DET
fcis-6912	6	21	named	name	VERB
fcis-6912	6	22	entity	entity	NOUN
fcis-6912	6	23	recognition	recognition	NOUN
fcis-6912	6	24	task	task	NOUN
fcis-6912	6	25	.	.	PUNCT
fcis-6912	7	1	the	the	DET
fcis-6912	7	2	experimental	experimental	ADJ
fcis-6912	7	3	results	result	NOUN
fcis-6912	7	4	show	show	VERB
fcis-6912	7	5	that	that	SCONJ
fcis-6912	7	6	compared	compare	VERB
fcis-6912	7	7	with	with	ADP
fcis-6912	7	8	the	the	DET
fcis-6912	7	9	traditional	traditional	ADJ
fcis-6912	7	10	model	model	NOUN
fcis-6912	7	11	,	,	PUNCT
fcis-6912	7	12	the	the	DET
fcis-6912	7	13	f1	f1	ADJ
fcis-6912	7	14	score	score	NOUN
fcis-6912	7	15	of	of	ADP
fcis-6912	7	16	this	this	DET
fcis-6912	7	17	model	model	NOUN
fcis-6912	7	18	has	have	AUX
fcis-6912	7	19	significantly	significantly	ADV
fcis-6912	7	20	improved	improve	VERB
fcis-6912	7	21	.	.	PUNCT
fcis-6912	8	1	keywords	keyword	NOUN
fcis-6912	8	2	:	:	PUNCT
fcis-6912	8	3	named	name	VERB
fcis-6912	8	4	entity	entity	NOUN
fcis-6912	8	5	recognition	recognition	NOUN
fcis-6912	8	6	;	;	PUNCT
fcis-6912	8	7	enhanced	enhance	VERB
fcis-6912	8	8	representation	representation	NOUN
fcis-6912	8	9	through	through	ADP
fcis-6912	8	10	knowledge	knowledge	NOUN
fcis-6912	8	11	integration	integration	NOUN
fcis-6912	8	12	;	;	PUNCT
fcis-6912	8	13	gated	gate	VERB
fcis-6912	8	14	recurrent	recurrent	ADJ
fcis-6912	8	15	unit	unit	NOUN
fcis-6912	8	16	;	;	PUNCT
fcis-6912	8	17	conditional	conditional	ADJ
fcis-6912	8	18	random	random	ADJ
fcis-6912	8	19	field	field	NOUN
fcis-6912	8	20	.	.	PUNCT
fcis-6912	9	1	1	1	X
fcis-6912	9	2	.	.	X
fcis-6912	9	3	introduction	introduction	NOUN
fcis-6912	9	4	named	name	VERB
fcis-6912	9	5	entity	entity	NOUN
fcis-6912	9	6	recognition	recognition	NOUN
fcis-6912	9	7	(	(	PUNCT
fcis-6912	9	8	ner	ner	NOUN
fcis-6912	9	9	)	)	PUNCT
fcis-6912	10	1	[	[	X
fcis-6912	10	2	1]is	1]is	NUM
fcis-6912	10	3	a	a	DET
fcis-6912	10	4	fundamental	fundamental	ADJ
fcis-6912	10	5	task	task	NOUN
fcis-6912	10	6	in	in	ADP
fcis-6912	10	7	natural	natural	ADJ
fcis-6912	10	8	language	language	NOUN
fcis-6912	10	9	processing	processing	NOUN
fcis-6912	10	10	,	,	PUNCT
fcis-6912	10	11	aimed	aim	VERB
fcis-6912	10	12	at	at	ADP
fcis-6912	10	13	identifying	identify	VERB
fcis-6912	10	14	entities	entity	NOUN
fcis-6912	10	15	with	with	ADP
fcis-6912	10	16	specific	specific	ADJ
fcis-6912	10	17	meanings	meaning	NOUN
fcis-6912	10	18	,	,	PUNCT
fcis-6912	10	19	such	such	ADJ
fcis-6912	10	20	as	as	ADP
fcis-6912	10	21	names	name	NOUN
fcis-6912	10	22	of	of	ADP
fcis-6912	10	23	people	people	NOUN
fcis-6912	10	24	,	,	PUNCT
fcis-6912	10	25	places	place	NOUN
fcis-6912	10	26	,	,	PUNCT
fcis-6912	10	27	and	and	CCONJ
fcis-6912	10	28	organizations	organization	NOUN
fcis-6912	10	29	,	,	PUNCT
fcis-6912	10	30	from	from	ADP
fcis-6912	10	31	unstructured	unstructured	ADJ
fcis-6912	10	32	text	text	NOUN
fcis-6912	10	33	data	datum	NOUN
fcis-6912	10	34	.	.	PUNCT
fcis-6912	11	1	this	this	PRON
fcis-6912	11	2	provides	provide	VERB
fcis-6912	11	3	a	a	DET
fcis-6912	11	4	foundation	foundation	NOUN
fcis-6912	11	5	for	for	ADP
fcis-6912	11	6	subsequent	subsequent	ADJ
fcis-6912	11	7	applications	application	NOUN
fcis-6912	11	8	such	such	ADJ
fcis-6912	11	9	as	as	ADP
fcis-6912	11	10	information	information	NOUN
fcis-6912	11	11	extraction	extraction	NOUN
fcis-6912	11	12	,	,	PUNCT
fcis-6912	11	13	question	question	NOUN
fcis-6912	11	14	-	-	PUNCT
fcis-6912	11	15	answering	answer	VERB
fcis-6912	11	16	systems	system	NOUN
fcis-6912	11	17	,	,	PUNCT
fcis-6912	11	18	and	and	CCONJ
fcis-6912	11	19	machine	machine	NOUN
fcis-6912	11	20	translation	translation	NOUN
fcis-6912	11	21	.	.	PUNCT
fcis-6912	12	1	chinese	chinese	ADJ
fcis-6912	12	2	entity	entity	NOUN
fcis-6912	12	3	structures	structure	NOUN
fcis-6912	12	4	are	be	AUX
fcis-6912	12	5	complex	complex	ADJ
fcis-6912	12	6	,	,	PUNCT
fcis-6912	12	7	diverse	diverse	ADJ
fcis-6912	12	8	in	in	ADP
fcis-6912	12	9	form	form	NOUN
fcis-6912	12	10	,	,	PUNCT
fcis-6912	12	11	and	and	CCONJ
fcis-6912	12	12	have	have	VERB
fcis-6912	12	13	fuzzy	fuzzy	ADJ
fcis-6912	12	14	boundaries	boundary	NOUN
fcis-6912	12	15	[	[	X
fcis-6912	12	16	2	2	NUM
fcis-6912	12	17	]	]	PUNCT
fcis-6912	12	18	.	.	PUNCT
fcis-6912	13	1	moreover	moreover	ADV
fcis-6912	13	2	,	,	PUNCT
fcis-6912	13	3	chinese	chinese	ADJ
fcis-6912	13	4	characters	character	NOUN
fcis-6912	13	5	can	can	AUX
fcis-6912	13	6	have	have	VERB
fcis-6912	13	7	multiple	multiple	ADJ
fcis-6912	13	8	meanings	meaning	NOUN
fcis-6912	13	9	or	or	CCONJ
fcis-6912	13	10	be	be	AUX
fcis-6912	13	11	polysemous	polysemous	ADJ
fcis-6912	13	12	in	in	ADP
fcis-6912	13	13	different	different	ADJ
fcis-6912	13	14	contexts	contexts	NOUN
fcis-6912	13	15	.	.	PUNCT
fcis-6912	14	1	these	these	DET
fcis-6912	14	2	factors	factor	NOUN
fcis-6912	14	3	increase	increase	VERB
fcis-6912	14	4	the	the	DET
fcis-6912	14	5	difficulty	difficulty	NOUN
fcis-6912	14	6	of	of	ADP
fcis-6912	14	7	chinese	chinese	ADJ
fcis-6912	14	8	ner	ner	NOUN
fcis-6912	14	9	,	,	PUNCT
fcis-6912	14	10	making	make	VERB
fcis-6912	14	11	it	it	PRON
fcis-6912	14	12	more	more	ADV
fcis-6912	14	13	valuable	valuable	ADJ
fcis-6912	14	14	for	for	ADP
fcis-6912	14	15	research	research	NOUN
fcis-6912	14	16	and	and	CCONJ
fcis-6912	14	17	practical	practical	ADJ
fcis-6912	14	18	applications	application	NOUN
fcis-6912	14	19	.	.	PUNCT
fcis-6912	15	1	the	the	DET
fcis-6912	15	2	ernie	ernie	PROPN
fcis-6912	15	3	-	-	PUNCT
fcis-6912	15	4	bilstm	bilstm	NOUN
fcis-6912	15	5	-	-	PUNCT
fcis-6912	15	6	crf	crf	NOUN
fcis-6912	15	7	model	model	NOUN
fcis-6912	15	8	is	be	AUX
fcis-6912	15	9	proposed	propose	VERB
fcis-6912	15	10	.	.	PUNCT
fcis-6912	16	1	the	the	DET
fcis-6912	16	2	ernie	ernie	PROPN
fcis-6912	16	3	pre	pre	ADJ
fcis-6912	16	4	-	-	ADJ
fcis-6912	16	5	training	training	ADJ
fcis-6912	16	6	model	model	NOUN
fcis-6912	16	7	can	can	AUX
fcis-6912	16	8	output	output	VERB
fcis-6912	16	9	different	different	ADJ
fcis-6912	16	10	word	word	NOUN
fcis-6912	16	11	vectors	vector	NOUN
fcis-6912	16	12	for	for	ADP
fcis-6912	16	13	different	different	ADJ
fcis-6912	16	14	contexts	context	NOUN
fcis-6912	16	15	by	by	ADP
fcis-6912	16	16	using	use	VERB
fcis-6912	16	17	multiple	multiple	ADJ
fcis-6912	16	18	layers	layer	NOUN
fcis-6912	16	19	of	of	ADP
fcis-6912	16	20	transformer	transformer	NOUN
fcis-6912	16	21	,	,	PUNCT
fcis-6912	16	22	obtaining	obtain	VERB
fcis-6912	16	23	dynamic	dynamic	ADJ
fcis-6912	16	24	word	word	NOUN
fcis-6912	16	25	vectors	vector	NOUN
fcis-6912	16	26	that	that	PRON
fcis-6912	16	27	contain	contain	VERB
fcis-6912	16	28	overall	overall	ADJ
fcis-6912	16	29	sequence	sequence	NOUN
fcis-6912	16	30	information	information	NOUN
fcis-6912	16	31	.	.	PUNCT
fcis-6912	17	1	secondly	secondly	ADV
fcis-6912	17	2	,	,	PUNCT
fcis-6912	17	3	the	the	DET
fcis-6912	17	4	word	word	NOUN
fcis-6912	17	5	vectors	vector	NOUN
fcis-6912	17	6	are	be	AUX
fcis-6912	17	7	input	input	VERB
fcis-6912	17	8	into	into	ADP
fcis-6912	17	9	the	the	DET
fcis-6912	17	10	bilstm	bilstm	NOUN
fcis-6912	17	11	layer	layer	NOUN
fcis-6912	17	12	,	,	PUNCT
fcis-6912	17	13	which	which	PRON
fcis-6912	17	14	can	can	AUX
fcis-6912	17	15	obtain	obtain	VERB
fcis-6912	17	16	sentence	sentence	NOUN
fcis-6912	17	17	context	context	NOUN
fcis-6912	17	18	information	information	NOUN
fcis-6912	17	19	through	through	ADP
fcis-6912	17	20	forward	forward	ADV
fcis-6912	17	21	and	and	CCONJ
fcis-6912	17	22	backward	backward	ADJ
fcis-6912	17	23	lstm	lstm	NOUN
fcis-6912	17	24	and	and	CCONJ
fcis-6912	17	25	obtain	obtain	VERB
fcis-6912	17	26	more	more	ADJ
fcis-6912	17	27	sentence	sentence	NOUN
fcis-6912	17	28	features	feature	NOUN
fcis-6912	17	29	,	,	PUNCT
fcis-6912	17	30	thereby	thereby	ADV
fcis-6912	17	31	improving	improve	VERB
fcis-6912	17	32	the	the	DET
fcis-6912	17	33	model	model	NOUN
fcis-6912	17	34	's	's	PART
fcis-6912	17	35	effectiveness	effectiveness	NOUN
fcis-6912	17	36	.	.	PUNCT
fcis-6912	18	1	finally	finally	ADV
fcis-6912	18	2	,	,	PUNCT
fcis-6912	18	3	the	the	DET
fcis-6912	18	4	sequence	sequence	NOUN
fcis-6912	18	5	is	be	AUX
fcis-6912	18	6	labeled	label	VERB
fcis-6912	18	7	through	through	ADP
fcis-6912	18	8	the	the	DET
fcis-6912	18	9	crf	crf	NOUN
fcis-6912	18	10	layer	layer	NOUN
fcis-6912	18	11	to	to	PART
fcis-6912	18	12	obtain	obtain	VERB
fcis-6912	18	13	the	the	DET
fcis-6912	18	14	globally	globally	ADV
fcis-6912	18	15	optimal	optimal	ADJ
fcis-6912	18	16	labeling	labeling	NOUN
fcis-6912	18	17	information	information	NOUN
fcis-6912	18	18	and	and	CCONJ
fcis-6912	18	19	complete	complete	VERB
fcis-6912	18	20	the	the	DET
fcis-6912	18	21	named	name	VERB
fcis-6912	18	22	entity	entity	NOUN
fcis-6912	18	23	recognition	recognition	NOUN
fcis-6912	18	24	task	task	NOUN
fcis-6912	18	25	.	.	PUNCT
fcis-6912	19	1	the	the	DET
fcis-6912	19	2	experimental	experimental	ADJ
fcis-6912	19	3	results	result	NOUN
fcis-6912	19	4	show	show	VERB
fcis-6912	19	5	that	that	SCONJ
fcis-6912	19	6	compared	compare	VERB
fcis-6912	19	7	with	with	ADP
fcis-6912	19	8	the	the	DET
fcis-6912	19	9	traditional	traditional	ADJ
fcis-6912	19	10	model	model	NOUN
fcis-6912	19	11	,	,	PUNCT
fcis-6912	19	12	the	the	DET
fcis-6912	19	13	f1	f1	ADJ
fcis-6912	19	14	score	score	NOUN
fcis-6912	19	15	of	of	ADP
fcis-6912	19	16	this	this	DET
fcis-6912	19	17	model	model	NOUN
fcis-6912	19	18	has	have	AUX
fcis-6912	19	19	significantly	significantly	ADV
fcis-6912	19	20	improved	improve	VERB
fcis-6912	19	21	.	.	PUNCT
fcis-6912	20	1	2	2	X
fcis-6912	20	2	.	.	X
fcis-6912	20	3	related	relate	VERB
fcis-6912	20	4	work	work	NOUN
fcis-6912	20	5	named	name	VERB
fcis-6912	20	6	entity	entity	NOUN
fcis-6912	20	7	recognition	recognition	NOUN
fcis-6912	20	8	(	(	PUNCT
fcis-6912	20	9	ner	ner	NOUN
fcis-6912	20	10	)	)	PUNCT
fcis-6912	20	11	is	be	AUX
fcis-6912	20	12	the	the	DET
fcis-6912	20	13	extraction	extraction	NOUN
fcis-6912	20	14	of	of	ADP
fcis-6912	20	15	required	require	VERB
fcis-6912	20	16	entities	entity	NOUN
fcis-6912	20	17	from	from	ADP
fcis-6912	20	18	structured	structured	ADJ
fcis-6912	20	19	or	or	CCONJ
fcis-6912	20	20	unstructured	unstructured	ADJ
fcis-6912	20	21	text	text	NOUN
fcis-6912	20	22	.	.	PUNCT
fcis-6912	21	1	in	in	ADP
fcis-6912	21	2	1996	1996	NUM
fcis-6912	21	3	,	,	PUNCT
fcis-6912	21	4	the	the	DET
fcis-6912	21	5	muc-6	muc-6	NUM
fcis-6912	21	6	conference	conference	NOUN
fcis-6912	21	7	first	first	ADV
fcis-6912	21	8	proposed	propose	VERB
fcis-6912	21	9	the	the	DET
fcis-6912	21	10	task	task	NOUN
fcis-6912	21	11	of	of	ADP
fcis-6912	21	12	named	name	VERB
fcis-6912	21	13	entity	entity	NOUN
fcis-6912	21	14	recognition	recognition	NOUN
fcis-6912	21	15	and	and	CCONJ
fcis-6912	21	16	specified	specify	VERB
fcis-6912	21	17	the	the	DET
fcis-6912	21	18	main	main	ADJ
fcis-6912	21	19	task	task	NOUN
fcis-6912	21	20	as	as	ADP
fcis-6912	21	21	identifying	identify	VERB
fcis-6912	21	22	entities	entity	NOUN
fcis-6912	21	23	in	in	ADP
fcis-6912	21	24	the	the	DET
fcis-6912	21	25	text	text	NOUN
fcis-6912	21	26	to	to	PART
fcis-6912	21	27	be	be	AUX
fcis-6912	21	28	processed	process	VERB
fcis-6912	21	29	[	[	PUNCT
fcis-6912	21	30	3	3	NUM
fcis-6912	21	31	]	]	PUNCT
fcis-6912	21	32	.	.	PUNCT
fcis-6912	22	1	the	the	DET
fcis-6912	22	2	entities	entity	NOUN
fcis-6912	22	3	that	that	PRON
fcis-6912	22	4	needed	need	VERB
fcis-6912	22	5	to	to	PART
fcis-6912	22	6	be	be	AUX
fcis-6912	22	7	identified	identify	VERB
fcis-6912	22	8	were	be	AUX
fcis-6912	22	9	divided	divide	VERB
fcis-6912	22	10	into	into	ADP
fcis-6912	22	11	three	three	NUM
fcis-6912	22	12	major	major	ADJ
fcis-6912	22	13	categories	category	NOUN
fcis-6912	22	14	(	(	PUNCT
fcis-6912	22	15	entity	entity	NOUN
fcis-6912	22	16	,	,	PUNCT
fcis-6912	22	17	time	time	NOUN
fcis-6912	22	18	,	,	PUNCT
fcis-6912	22	19	and	and	CCONJ
fcis-6912	22	20	numerical	numerical	ADJ
fcis-6912	22	21	)	)	PUNCT
fcis-6912	22	22	and	and	CCONJ
fcis-6912	22	23	seven	seven	NUM
fcis-6912	22	24	subcategories	subcategorie	NOUN
fcis-6912	22	25	(	(	PUNCT
fcis-6912	22	26	person	person	NOUN
fcis-6912	22	27	,	,	PUNCT
fcis-6912	22	28	organization	organization	NOUN
fcis-6912	22	29	,	,	PUNCT
fcis-6912	22	30	location	location	NOUN
fcis-6912	22	31	,	,	PUNCT
fcis-6912	22	32	time	time	NOUN
fcis-6912	22	33	,	,	PUNCT
fcis-6912	22	34	date	date	NOUN
fcis-6912	22	35	,	,	PUNCT
fcis-6912	22	36	currency	currency	NOUN
fcis-6912	22	37	,	,	PUNCT
fcis-6912	22	38	and	and	CCONJ
fcis-6912	22	39	percentage	percentage	NOUN
fcis-6912	22	40	)	)	PUNCT
fcis-6912	22	41	.	.	PUNCT
fcis-6912	23	1	the	the	DET
fcis-6912	23	2	named	name	VERB
fcis-6912	23	3	entity	entity	NOUN
fcis-6912	23	4	recognition	recognition	NOUN
fcis-6912	23	5	task	task	NOUN
fcis-6912	23	6	proposed	propose	VERB
fcis-6912	23	7	by	by	ADP
fcis-6912	23	8	the	the	DET
fcis-6912	23	9	muc-6	muc-6	NUM
fcis-6912	23	10	conference	conference	NOUN
fcis-6912	23	11	has	have	AUX
fcis-6912	23	12	been	be	AUX
fcis-6912	23	13	of	of	ADP
fcis-6912	23	14	great	great	ADJ
fcis-6912	23	15	interest	interest	NOUN
fcis-6912	23	16	to	to	ADP
fcis-6912	23	17	people	people	NOUN
fcis-6912	23	18	,	,	PUNCT
fcis-6912	23	19	and	and	CCONJ
fcis-6912	23	20	subsequently	subsequently	ADV
fcis-6912	23	21	,	,	PUNCT
fcis-6912	23	22	conferences	conference	NOUN
fcis-6912	23	23	such	such	ADJ
fcis-6912	23	24	as	as	ADP
fcis-6912	23	25	ace	ace	NOUN
fcis-6912	23	26	and	and	CCONJ
fcis-6912	23	27	conll-2003	conll-2003	NOUN
fcis-6912	23	28	have	have	AUX
fcis-6912	23	29	also	also	ADV
fcis-6912	23	30	proposed	propose	VERB
fcis-6912	23	31	research	research	NOUN
fcis-6912	23	32	on	on	ADP
fcis-6912	23	33	named	name	VERB
fcis-6912	23	34	entity	entity	NOUN
fcis-6912	23	35	recognition	recognition	NOUN
fcis-6912	23	36	[	[	X
fcis-6912	23	37	4	4	NUM
fcis-6912	23	38	]	]	PUNCT
fcis-6912	23	39	.	.	PUNCT
fcis-6912	24	1	in	in	ADP
fcis-6912	24	2	these	these	DET
fcis-6912	24	3	conferences	conference	NOUN
fcis-6912	24	4	,	,	PUNCT
fcis-6912	24	5	not	not	PART
fcis-6912	24	6	only	only	ADV
fcis-6912	24	7	were	be	AUX
fcis-6912	24	8	the	the	DET
fcis-6912	24	9	entities	entity	NOUN
fcis-6912	24	10	specified	specify	VERB
fcis-6912	24	11	by	by	ADP
fcis-6912	24	12	muc-6	muc-6	SYM
fcis-6912	24	13	identified	identify	VERB
fcis-6912	24	14	,	,	PUNCT
fcis-6912	24	15	but	but	CCONJ
fcis-6912	24	16	new	new	ADJ
fcis-6912	24	17	entity	entity	NOUN
fcis-6912	24	18	types	type	NOUN
fcis-6912	24	19	were	be	AUX
fcis-6912	24	20	also	also	ADV
fcis-6912	24	21	added	add	VERB
fcis-6912	24	22	.	.	PUNCT
fcis-6912	25	1	for	for	ADP
fcis-6912	25	2	example	example	NOUN
fcis-6912	25	3	,	,	PUNCT
fcis-6912	25	4	the	the	DET
fcis-6912	25	5	ace	ace	PROPN
fcis-6912	25	6	conference	conference	NOUN
fcis-6912	25	7	added	add	VERB
fcis-6912	25	8	multiple	multiple	ADJ
fcis-6912	25	9	entity	entity	NOUN
fcis-6912	25	10	types	type	NOUN
fcis-6912	25	11	such	such	ADJ
fcis-6912	25	12	as	as	ADP
fcis-6912	25	13	geopolitical	geopolitical	ADJ
fcis-6912	25	14	,	,	PUNCT
fcis-6912	25	15	facility	facility	NOUN
fcis-6912	25	16	transportation	transportation	NOUN
fcis-6912	25	17	,	,	PUNCT
fcis-6912	25	18	and	and	CCONJ
fcis-6912	25	19	weapons	weapon	NOUN
fcis-6912	25	20	.	.	PUNCT
fcis-6912	26	1	additionally	additionally	ADV
fcis-6912	26	2	,	,	PUNCT
fcis-6912	26	3	ace	ace	VERB
fcis-6912	26	4	and	and	CCONJ
fcis-6912	26	5	conll-2003	conll-2003	NOUN
fcis-6912	26	6	introduced	introduce	VERB
fcis-6912	26	7	multiple	multiple	ADJ
fcis-6912	26	8	languages	language	NOUN
fcis-6912	26	9	in	in	ADP
fcis-6912	26	10	the	the	DET
fcis-6912	26	11	named	name	VERB
fcis-6912	26	12	entity	entity	NOUN
fcis-6912	26	13	recognition	recognition	NOUN
fcis-6912	26	14	task	task	NOUN
fcis-6912	26	15	,	,	PUNCT
fcis-6912	26	16	such	such	ADJ
fcis-6912	26	17	as	as	ADP
fcis-6912	26	18	spanish	spanish	ADJ
fcis-6912	26	19	and	and	CCONJ
fcis-6912	26	20	chinese	chinese	ADJ
fcis-6912	26	21	in	in	ADP
fcis-6912	26	22	ace	ace	NOUN
fcis-6912	26	23	,	,	PUNCT
fcis-6912	26	24	and	and	CCONJ
fcis-6912	26	25	english	english	PROPN
fcis-6912	26	26	and	and	CCONJ
fcis-6912	26	27	german	german	NOUN
fcis-6912	26	28	in	in	ADP
fcis-6912	26	29	conll-2003[5	conll-2003[5	PROPN
fcis-6912	26	30	]	]	PUNCT
fcis-6912	26	31	.	.	PUNCT
fcis-6912	27	1	moreover	moreover	ADV
fcis-6912	27	2	,	,	PUNCT
fcis-6912	27	3	there	there	PRON
fcis-6912	27	4	were	be	VERB
fcis-6912	27	5	differences	difference	NOUN
fcis-6912	27	6	in	in	ADP
fcis-6912	27	7	the	the	DET
fcis-6912	27	8	research	research	NOUN
fcis-6912	27	9	tasks	task	NOUN
fcis-6912	27	10	and	and	CCONJ
fcis-6912	27	11	mainstream	mainstream	NOUN
fcis-6912	27	12	methods	method	NOUN
fcis-6912	27	13	.	.	PUNCT
fcis-6912	28	1	ace	ace	VERB
fcis-6912	28	2	not	not	PART
fcis-6912	28	3	only	only	ADV
fcis-6912	28	4	carried	carry	VERB
fcis-6912	28	5	out	out	ADP
fcis-6912	28	6	the	the	DET
fcis-6912	28	7	task	task	NOUN
fcis-6912	28	8	of	of	ADP
fcis-6912	28	9	named	name	VERB
fcis-6912	28	10	entity	entity	NOUN
fcis-6912	28	11	recognition	recognition	NOUN
fcis-6912	28	12	but	but	CCONJ
fcis-6912	28	13	also	also	ADV
fcis-6912	28	14	included	include	VERB
fcis-6912	28	15	relationship	relationship	NOUN
fcis-6912	28	16	recognition	recognition	NOUN
fcis-6912	28	17	,	,	PUNCT
fcis-6912	28	18	reference	reference	NOUN
fcis-6912	28	19	,	,	PUNCT
fcis-6912	28	20	and	and	CCONJ
fcis-6912	28	21	anaphora	anaphora	ADJ
fcis-6912	28	22	resolution	resolution	NOUN
fcis-6912	28	23	as	as	ADP
fcis-6912	28	24	part	part	NOUN
fcis-6912	28	25	of	of	ADP
fcis-6912	28	26	the	the	DET
fcis-6912	28	27	research	research	NOUN
fcis-6912	28	28	[	[	X
fcis-6912	28	29	6	6	NUM
fcis-6912	28	30	]	]	PUNCT
fcis-6912	28	31	.	.	PUNCT
fcis-6912	29	1	in	in	ADP
fcis-6912	29	2	contrast	contrast	NOUN
fcis-6912	29	3	to	to	ADP
fcis-6912	29	4	muc-6	muc-6	PRON
fcis-6912	29	5	,	,	PUNCT
fcis-6912	29	6	the	the	DET
fcis-6912	29	7	methods	method	NOUN
fcis-6912	29	8	that	that	PRON
fcis-6912	29	9	performed	perform	VERB
fcis-6912	29	10	well	well	ADV
fcis-6912	29	11	in	in	ADP
fcis-6912	29	12	conll-2003	conll-2003	NOUN
fcis-6912	29	13	directly	directly	ADV
fcis-6912	29	14	or	or	CCONJ
fcis-6912	29	15	indirectly	indirectly	ADV
fcis-6912	29	16	used	use	VERB
fcis-6912	29	17	the	the	DET
fcis-6912	29	18	maximum	maximum	PROPN
fcis-6912	29	19	entropy	entropy	NOUN
fcis-6912	29	20	model	model	NOUN
fcis-6912	29	21	or	or	CCONJ
fcis-6912	29	22	some	some	DET
fcis-6912	29	23	statistical	statistical	ADJ
fcis-6912	29	24	machine	machine	NOUN
fcis-6912	29	25	learning	learning	NOUN
fcis-6912	29	26	methods	method	NOUN
fcis-6912	29	27	,	,	PUNCT
fcis-6912	29	28	while	while	SCONJ
fcis-6912	29	29	in	in	ADP
fcis-6912	29	30	muc-6	muc-6	NUM
fcis-6912	29	31	,	,	PUNCT
fcis-6912	29	32	rule	rule	NOUN
fcis-6912	29	33	-	-	PUNCT
fcis-6912	29	34	based	base	VERB
fcis-6912	29	35	and	and	CCONJ
fcis-6912	29	36	dictionary	dictionary	ADJ
fcis-6912	29	37	-	-	PUNCT
fcis-6912	29	38	based	base	VERB
fcis-6912	29	39	methods	method	NOUN
fcis-6912	29	40	were	be	AUX
fcis-6912	29	41	the	the	DET
fcis-6912	29	42	mainstream	mainstream	NOUN
fcis-6912	29	43	methods	method	NOUN
fcis-6912	29	44	at	at	ADP
fcis-6912	29	45	the	the	DET
fcis-6912	29	46	time	time	NOUN
fcis-6912	29	47	[	[	X
fcis-6912	29	48	7	7	NUM
fcis-6912	29	49	]	]	PUNCT
fcis-6912	29	50	.	.	PUNCT
fcis-6912	30	1	bootstrapping	bootstrappe	VERB
fcis-6912	30	2	[	[	X
fcis-6912	30	3	8	8	NUM
fcis-6912	30	4	]	]	PUNCT
fcis-6912	30	5	is	be	AUX
fcis-6912	30	6	a	a	DET
fcis-6912	30	7	classic	classic	ADJ
fcis-6912	30	8	method	method	NOUN
fcis-6912	30	9	that	that	PRON
fcis-6912	30	10	can	can	AUX
fcis-6912	30	11	automatically	automatically	ADV
fcis-6912	30	12	generate	generate	VERB
fcis-6912	30	13	rules	rule	NOUN
fcis-6912	30	14	.	.	PUNCT
fcis-6912	31	1	the	the	DET
fcis-6912	31	2	rule	rule	NOUN
fcis-6912	31	3	-	-	PUNCT
fcis-6912	31	4	based	base	VERB
fcis-6912	31	5	and	and	CCONJ
fcis-6912	31	6	dictionarybased	dictionarybase	VERB
fcis-6912	31	7	methods	method	NOUN
fcis-6912	31	8	require	require	VERB
fcis-6912	31	9	manual	manual	ADJ
fcis-6912	31	10	customization	customization	NOUN
fcis-6912	31	11	by	by	ADP
fcis-6912	31	12	relevant	relevant	ADJ
fcis-6912	31	13	personnel	personnel	NOUN
fcis-6912	31	14	.	.	PUNCT
fcis-6912	32	1	in	in	ADP
fcis-6912	32	2	the	the	DET
fcis-6912	32	3	current	current	ADJ
fcis-6912	32	4	field	field	NOUN
fcis-6912	32	5	and	and	CCONJ
fcis-6912	32	6	scope	scope	NOUN
fcis-6912	32	7	of	of	ADP
fcis-6912	32	8	corpora	corpus	NOUN
fcis-6912	32	9	,	,	PUNCT
fcis-6912	32	10	these	these	DET
fcis-6912	32	11	methods	method	NOUN
fcis-6912	32	12	need	need	VERB
fcis-6912	32	13	to	to	PART
fcis-6912	32	14	gradually	gradually	ADV
fcis-6912	32	15	increase	increase	VERB
fcis-6912	32	16	in	in	ADP
fcis-6912	32	17	size	size	NOUN
fcis-6912	32	18	to	to	PART
fcis-6912	32	19	adapt	adapt	VERB
fcis-6912	32	20	to	to	ADP
fcis-6912	32	21	new	new	ADJ
fcis-6912	32	22	situations	situation	NOUN
fcis-6912	32	23	,	,	PUNCT
fcis-6912	32	24	resulting	result	VERB
fcis-6912	32	25	in	in	ADP
fcis-6912	32	26	significantly	significantly	ADV
fcis-6912	32	27	increased	increase	VERB
fcis-6912	32	28	workload	workload	NOUN
fcis-6912	32	29	.	.	PUNCT
fcis-6912	33	1	therefore	therefore	ADV
fcis-6912	33	2	,	,	PUNCT
fcis-6912	33	3	rule	rule	NOUN
fcis-6912	33	4	-	-	PUNCT
fcis-6912	33	5	based	base	VERB
fcis-6912	33	6	and	and	CCONJ
fcis-6912	33	7	dictionary	dictionary	ADJ
fcis-6912	33	8	-	-	PUNCT
fcis-6912	33	9	based	base	VERB
fcis-6912	33	10	methods	method	NOUN
fcis-6912	33	11	are	be	AUX
fcis-6912	33	12	rarely	rarely	ADV
fcis-6912	33	13	used	use	VERB
fcis-6912	33	14	alone	alone	ADV
fcis-6912	33	15	nowadays	nowadays	ADV
fcis-6912	33	16	but	but	CCONJ
fcis-6912	33	17	are	be	AUX
fcis-6912	33	18	mainly	mainly	ADV
fcis-6912	33	19	used	use	VERB
fcis-6912	33	20	in	in	ADP
fcis-6912	33	21	combination	combination	NOUN
fcis-6912	33	22	with	with	ADP
fcis-6912	33	23	other	other	ADJ
fcis-6912	33	24	methods	method	NOUN
fcis-6912	33	25	to	to	PART
fcis-6912	33	26	improve	improve	VERB
fcis-6912	33	27	model	model	NOUN
fcis-6912	33	28	accuracy	accuracy	NOUN
fcis-6912	33	29	.	.	PUNCT
fcis-6912	34	1	after	after	ADP
fcis-6912	34	2	the	the	DET
fcis-6912	34	3	use	use	NOUN
fcis-6912	34	4	of	of	ADP
fcis-6912	34	5	rule	rule	NOUN
fcis-6912	34	6	-	-	PUNCT
fcis-6912	34	7	based	base	VERB
fcis-6912	34	8	and	and	CCONJ
fcis-6912	34	9	dictionary	dictionary	ADJ
fcis-6912	34	10	-	-	PUNCT
fcis-6912	34	11	based	base	VERB
fcis-6912	34	12	methods	method	NOUN
fcis-6912	34	13	,	,	PUNCT
fcis-6912	34	14	people	people	NOUN
fcis-6912	34	15	began	begin	VERB
fcis-6912	34	16	to	to	PART
fcis-6912	34	17	use	use	VERB
fcis-6912	34	18	statistical	statistical	ADJ
fcis-6912	34	19	learning	learning	NOUN
fcis-6912	34	20	methods	method	NOUN
fcis-6912	34	21	for	for	ADP
fcis-6912	34	22	named	name	VERB
fcis-6912	34	23	entity	entity	NOUN
fcis-6912	34	24	recognition	recognition	NOUN
fcis-6912	34	25	research	research	NOUN
fcis-6912	34	26	.	.	PUNCT
fcis-6912	35	1	the	the	DET
fcis-6912	35	2	conditional	conditional	ADJ
fcis-6912	35	3	random	random	ADJ
fcis-6912	35	4	fields	field	NOUN
fcis-6912	35	5	(	(	PUNCT
fcis-6912	35	6	crf	crf	NOUN
fcis-6912	35	7	)	)	PUNCT
fcis-6912	35	8	model	model	NOUN
fcis-6912	35	9	was	be	AUX
fcis-6912	35	10	used	use	VERB
fcis-6912	35	11	later	later	ADV
fcis-6912	35	12	than	than	ADP
fcis-6912	35	13	the	the	DET
fcis-6912	35	14	support	support	NOUN
fcis-6912	35	15	vector	vector	NOUN
fcis-6912	35	16	machine	machine	NOUN
fcis-6912	35	17	(	(	PUNCT
fcis-6912	35	18	svm	svm	PROPN
fcis-6912	35	19	)	)	PUNCT
fcis-6912	35	20	model	model	NOUN
fcis-6912	35	21	,	,	PUNCT
fcis-6912	35	22	but	but	CCONJ
fcis-6912	35	23	it	it	PRON
fcis-6912	35	24	was	be	AUX
fcis-6912	35	25	more	more	ADV
fcis-6912	35	26	effective	effective	ADJ
fcis-6912	35	27	,	,	PUNCT
fcis-6912	35	28	and	and	CCONJ
fcis-6912	35	29	it	it	PRON
fcis-6912	35	30	is	be	AUX
fcis-6912	35	31	now	now	ADV
fcis-6912	35	32	one	one	NUM
fcis-6912	35	33	of	of	ADP
fcis-6912	35	34	the	the	DET
fcis-6912	35	35	most	most	ADV
fcis-6912	35	36	commonly	commonly	ADV
fcis-6912	35	37	used	use	VERB
fcis-6912	35	38	models	model	NOUN
fcis-6912	35	39	.	.	PUNCT
fcis-6912	36	1	researchers	researcher	NOUN
fcis-6912	36	2	such	such	ADJ
fcis-6912	36	3	as	as	ADP
fcis-6912	36	4	li[9	li[9	PROPN
fcis-6912	36	5	]	]	PUNCT
fcis-6912	36	6	and	and	CCONJ
fcis-6912	36	7	jiang[10	jiang[10	PROPN
fcis-6912	36	8	]	]	PUNCT
fcis-6912	36	9	have	have	AUX
fcis-6912	36	10	conducted	conduct	VERB
fcis-6912	36	11	a	a	DET
fcis-6912	36	12	series	series	NOUN
fcis-6912	36	13	of	of	ADP
fcis-6912	36	14	model	model	PROPN
fcis-6912	36	15	comparison	comparison	NOUN
fcis-6912	36	16	studies	study	NOUN
fcis-6912	36	17	,	,	PUNCT
fcis-6912	36	18	and	and	CCONJ
fcis-6912	36	19	from	from	ADP
fcis-6912	36	20	the	the	DET
fcis-6912	36	21	experimental	experimental	ADJ
fcis-6912	36	22	results	result	NOUN
fcis-6912	36	23	,	,	PUNCT
fcis-6912	36	24	it	it	PRON
fcis-6912	36	25	can	can	AUX
fcis-6912	36	26	be	be	AUX
fcis-6912	36	27	concluded	conclude	VERB
fcis-6912	36	28	that	that	SCONJ
fcis-6912	36	29	crf	crf	PROPN
fcis-6912	36	30	is	be	AUX
fcis-6912	36	31	a	a	DET
fcis-6912	36	32	better	well	ADJ
fcis-6912	36	33	model	model	NOUN
fcis-6912	36	34	for	for	ADP
fcis-6912	36	35	named	name	VERB
fcis-6912	36	36	entity	entity	NOUN
fcis-6912	36	37	recognition	recognition	NOUN
fcis-6912	36	38	.	.	PUNCT
fcis-6912	37	1	stanford	stanford	PROPN
fcis-6912	37	2	university	university	PROPN
fcis-6912	37	3	has	have	AUX
fcis-6912	37	4	also	also	ADV
fcis-6912	37	5	developed	develop	VERB
fcis-6912	37	6	the	the	DET
fcis-6912	37	7	standorfner	standorfner	ADJ
fcis-6912	37	8	tool	tool	NOUN
fcis-6912	37	9	for	for	ADP
fcis-6912	37	10	named	name	VERB
fcis-6912	37	11	entity	entity	NOUN
fcis-6912	37	12	recognition	recognition	NOUN
fcis-6912	37	13	based	base	VERB
fcis-6912	37	14	on	on	ADP
fcis-6912	37	15	crf	crf	PROPN
fcis-6912	37	16	.	.	PUNCT
fcis-6912	37	17	multi	multi	ADJ
fcis-6912	37	18	-	-	ADJ
fcis-6912	37	19	model	model	ADJ
fcis-6912	37	20	mixing	mixing	NOUN
fcis-6912	37	21	is	be	AUX
fcis-6912	37	22	a	a	DET
fcis-6912	37	23	choice	choice	NOUN
fcis-6912	37	24	for	for	ADP
fcis-6912	37	25	improving	improve	VERB
fcis-6912	37	26	model	model	NOUN
fcis-6912	37	27	performance	performance	NOUN
fcis-6912	37	28	.	.	PUNCT
fcis-6912	38	1	li[11	li[11	X
fcis-6912	38	2	]	]	PUNCT
fcis-6912	38	3	and	and	CCONJ
fcis-6912	38	4	others	other	NOUN
fcis-6912	38	5	used	use	VERB
fcis-6912	38	6	multiple	multiple	ADJ
fcis-6912	38	7	svm	svm	ADJ
fcis-6912	38	8	models	model	NOUN
fcis-6912	38	9	to	to	PART
fcis-6912	38	10	improve	improve	VERB
fcis-6912	38	11	model	model	NOUN
fcis-6912	38	12	performance	performance	NOUN
fcis-6912	38	13	.	.	PUNCT
fcis-6912	39	1	the	the	DET
fcis-6912	39	2	use	use	NOUN
fcis-6912	39	3	of	of	ADP
fcis-6912	39	4	external	external	ADJ
fcis-6912	39	5	knowledge	knowledge	NOUN
fcis-6912	39	6	bases	basis	NOUN
fcis-6912	39	7	to	to	PART
fcis-6912	39	8	solve	solve	VERB
fcis-6912	39	9	new	new	ADJ
fcis-6912	39	10	entities	entity	NOUN
fcis-6912	39	11	is	be	AUX
fcis-6912	39	12	a	a	DET
fcis-6912	39	13	method	method	NOUN
fcis-6912	39	14	chosen	choose	VERB
fcis-6912	39	15	by	by	ADP
fcis-6912	39	16	many	many	ADJ
fcis-6912	39	17	researchers	researcher	NOUN
fcis-6912	39	18	,	,	PUNCT
fcis-6912	39	19	such	such	ADJ
fcis-6912	39	20	as	as	ADP
fcis-6912	39	21	cukierman	cukierman	NOUN
fcis-6912	39	22	[	[	X
fcis-6912	39	23	12	12	NUM
fcis-6912	39	24	]	]	PUNCT
fcis-6912	39	25	using	use	VERB
fcis-6912	39	26	the	the	DET
fcis-6912	39	27	wikipedia	wikipedia	PROPN
fcis-6912	39	28	database	database	NOUN
fcis-6912	39	29	for	for	ADP
fcis-6912	39	30	semantic	semantic	ADJ
fcis-6912	39	31	disambiguation	disambiguation	NOUN
fcis-6912	39	32	.	.	PUNCT
fcis-6912	40	1	22	22	NUM
fcis-6912	40	2	because	because	SCONJ
fcis-6912	40	3	traditional	traditional	ADJ
fcis-6912	40	4	machine	machine	NOUN
fcis-6912	40	5	learning	learning	NOUN
fcis-6912	40	6	models	model	NOUN
fcis-6912	40	7	heavily	heavily	ADV
fcis-6912	40	8	rely	rely	VERB
fcis-6912	40	9	on	on	ADP
fcis-6912	40	10	feature	feature	NOUN
fcis-6912	40	11	engineering	engineering	NOUN
fcis-6912	40	12	,	,	PUNCT
fcis-6912	40	13	the	the	DET
fcis-6912	40	14	selection	selection	NOUN
fcis-6912	40	15	of	of	ADP
fcis-6912	40	16	feature	feature	NOUN
fcis-6912	40	17	engineering	engineering	NOUN
fcis-6912	40	18	is	be	AUX
fcis-6912	40	19	very	very	ADV
fcis-6912	40	20	complex	complex	ADJ
fcis-6912	40	21	.	.	PUNCT
fcis-6912	41	1	in	in	ADP
fcis-6912	41	2	recent	recent	ADJ
fcis-6912	41	3	years	year	NOUN
fcis-6912	41	4	,	,	PUNCT
fcis-6912	41	5	with	with	ADP
fcis-6912	41	6	the	the	DET
fcis-6912	41	7	rise	rise	NOUN
fcis-6912	41	8	of	of	ADP
fcis-6912	41	9	deep	deep	ADJ
fcis-6912	41	10	learning	learning	NOUN
fcis-6912	41	11	,	,	PUNCT
fcis-6912	41	12	representation	representation	NOUN
fcis-6912	41	13	learning	learning	NOUN
fcis-6912	41	14	has	have	AUX
fcis-6912	41	15	become	become	VERB
fcis-6912	41	16	a	a	DET
fcis-6912	41	17	research	research	NOUN
fcis-6912	41	18	hotspot	hotspot	NOUN
fcis-6912	41	19	.	.	PUNCT
fcis-6912	42	1	the	the	DET
fcis-6912	42	2	method	method	NOUN
fcis-6912	42	3	of	of	ADP
fcis-6912	42	4	word	word	NOUN
fcis-6912	42	5	vector	vector	NOUN
fcis-6912	42	6	representation	representation	NOUN
fcis-6912	42	7	in	in	ADP
fcis-6912	42	8	deep	deep	ADJ
fcis-6912	42	9	learning	learning	NOUN
fcis-6912	42	10	solves	solve	VERB
fcis-6912	42	11	the	the	DET
fcis-6912	42	12	problem	problem	NOUN
fcis-6912	42	13	of	of	ADP
fcis-6912	42	14	data	datum	NOUN
fcis-6912	42	15	sparsity	sparsity	NOUN
fcis-6912	42	16	,	,	PUNCT
fcis-6912	42	17	and	and	CCONJ
fcis-6912	42	18	word	word	NOUN
fcis-6912	42	19	vectors	vector	NOUN
fcis-6912	42	20	contain	contain	VERB
fcis-6912	42	21	semantic	semantic	ADJ
fcis-6912	42	22	information	information	NOUN
fcis-6912	42	23	.	.	PUNCT
fcis-6912	43	1	compared	compare	VERB
fcis-6912	43	2	with	with	ADP
fcis-6912	43	3	the	the	DET
fcis-6912	43	4	cumbersome	cumbersome	ADJ
fcis-6912	43	5	manual	manual	ADJ
fcis-6912	43	6	selection	selection	NOUN
fcis-6912	43	7	of	of	ADP
fcis-6912	43	8	features	feature	NOUN
fcis-6912	43	9	,	,	PUNCT
fcis-6912	43	10	using	use	VERB
fcis-6912	43	11	word	word	NOUN
fcis-6912	43	12	vectors	vector	NOUN
fcis-6912	43	13	is	be	AUX
fcis-6912	43	14	more	more	ADV
fcis-6912	43	15	suitable	suitable	ADJ
fcis-6912	43	16	for	for	ADP
fcis-6912	43	17	the	the	DET
fcis-6912	43	18	development	development	NOUN
fcis-6912	43	19	of	of	ADP
fcis-6912	43	20	named	name	VERB
fcis-6912	43	21	entity	entity	NOUN
fcis-6912	43	22	recognition	recognition	NOUN
fcis-6912	43	23	.	.	PUNCT
fcis-6912	44	1	under	under	ADP
fcis-6912	44	2	the	the	DET
fcis-6912	44	3	trend	trend	NOUN
fcis-6912	44	4	of	of	ADP
fcis-6912	44	5	deep	deep	ADJ
fcis-6912	44	6	learning	learning	NOUN
fcis-6912	44	7	,	,	PUNCT
fcis-6912	44	8	various	various	ADJ
fcis-6912	44	9	neural	neural	ADJ
fcis-6912	44	10	network	network	NOUN
fcis-6912	44	11	models	model	NOUN
fcis-6912	44	12	have	have	AUX
fcis-6912	44	13	been	be	AUX
fcis-6912	44	14	proposed	propose	VERB
fcis-6912	44	15	and	and	CCONJ
fcis-6912	44	16	improved	improve	VERB
fcis-6912	44	17	,	,	PUNCT
fcis-6912	44	18	and	and	CCONJ
fcis-6912	44	19	have	have	AUX
fcis-6912	44	20	been	be	AUX
fcis-6912	44	21	used	use	VERB
fcis-6912	44	22	in	in	ADP
fcis-6912	44	23	named	name	VERB
fcis-6912	44	24	entity	entity	NOUN
fcis-6912	44	25	recognition	recognition	NOUN
fcis-6912	44	26	tasks	task	NOUN
fcis-6912	44	27	,	,	PUNCT
fcis-6912	44	28	such	such	ADJ
fcis-6912	44	29	as	as	ADP
fcis-6912	44	30	convolutional	convolutional	ADJ
fcis-6912	44	31	neural	neural	ADJ
fcis-6912	44	32	network	network	NOUN
fcis-6912	44	33	(	(	PUNCT
fcis-6912	44	34	cnn	cnn	PROPN
fcis-6912	44	35	)	)	PUNCT
fcis-6912	44	36	,	,	PUNCT
fcis-6912	44	37	long	long	ADJ
fcis-6912	44	38	short	short	ADJ
fcis-6912	44	39	-	-	PUNCT
fcis-6912	44	40	term	term	NOUN
fcis-6912	44	41	memory	memory	NOUN
fcis-6912	44	42	network	network	NOUN
fcis-6912	44	43	(	(	PUNCT
fcis-6912	44	44	lstm	lstm	PROPN
fcis-6912	44	45	)	)	PUNCT
fcis-6912	44	46	,	,	PUNCT
fcis-6912	44	47	and	and	CCONJ
fcis-6912	44	48	bi	bi	ADJ
fcis-6912	44	49	-	-	ADJ
fcis-6912	44	50	directional	directional	ADJ
fcis-6912	44	51	long	long	ADJ
fcis-6912	44	52	short	short	ADJ
fcis-6912	44	53	-	-	PUNCT
fcis-6912	44	54	term	term	NOUN
fcis-6912	44	55	memory	memory	NOUN
fcis-6912	44	56	network	network	NOUN
fcis-6912	44	57	(	(	PUNCT
fcis-6912	44	58	bilstm	bilstm	NOUN
fcis-6912	44	59	)	)	PUNCT
fcis-6912	44	60	.	.	PUNCT
fcis-6912	45	1	in	in	ADP
fcis-6912	45	2	recent	recent	ADJ
fcis-6912	45	3	years	year	NOUN
fcis-6912	45	4	,	,	PUNCT
fcis-6912	45	5	the	the	DET
fcis-6912	45	6	bilstm	bilstm	NOUN
fcis-6912	45	7	-	-	PUNCT
fcis-6912	45	8	crf	crf	NOUN
fcis-6912	45	9	model	model	NOUN
fcis-6912	45	10	based	base	VERB
fcis-6912	45	11	on	on	ADP
fcis-6912	45	12	deep	deep	ADJ
fcis-6912	45	13	learning	learning	NOUN
fcis-6912	45	14	has	have	AUX
fcis-6912	45	15	been	be	AUX
fcis-6912	45	16	more	more	ADV
fcis-6912	45	17	effective	effective	ADJ
fcis-6912	45	18	in	in	ADP
fcis-6912	45	19	medical	medical	ADJ
fcis-6912	45	20	named	name	VERB
fcis-6912	45	21	entity	entity	NOUN
fcis-6912	45	22	recognition	recognition	NOUN
fcis-6912	45	23	.	.	PUNCT
fcis-6912	46	1	collobert	collobert	AUX
fcis-6912	46	2	et	et	PROPN
fcis-6912	46	3	al	al	PROPN
fcis-6912	46	4	.	.	PROPN
fcis-6912	46	5	proposed	propose	VERB
fcis-6912	46	6	the	the	DET
fcis-6912	46	7	most	most	ADV
fcis-6912	46	8	representative	representative	ADJ
fcis-6912	46	9	deep	deep	ADJ
fcis-6912	46	10	learning	learning	NOUN
fcis-6912	46	11	model	model	NOUN
fcis-6912	46	12	and	and	CCONJ
fcis-6912	46	13	designed	design	VERB
fcis-6912	46	14	the	the	DET
fcis-6912	46	15	senna	senna	NOUN
fcis-6912	46	16	[	[	X
fcis-6912	46	17	13	13	NUM
fcis-6912	46	18	]	]	PUNCT
fcis-6912	46	19	system	system	NOUN
fcis-6912	46	20	,	,	PUNCT
fcis-6912	46	21	which	which	PRON
fcis-6912	46	22	requires	require	VERB
fcis-6912	46	23	small	small	ADJ
fcis-6912	46	24	memory	memory	NOUN
fcis-6912	46	25	and	and	CCONJ
fcis-6912	46	26	can	can	AUX
fcis-6912	46	27	efficiently	efficiently	ADV
fcis-6912	46	28	solve	solve	VERB
fcis-6912	46	29	various	various	ADJ
fcis-6912	46	30	problems	problem	NOUN
fcis-6912	46	31	such	such	ADJ
fcis-6912	46	32	as	as	ADP
fcis-6912	46	33	part	part	NOUN
fcis-6912	46	34	-	-	PUNCT
fcis-6912	46	35	of	of	ADP
fcis-6912	46	36	-	-	PUNCT
fcis-6912	46	37	speech	speech	NOUN
fcis-6912	46	38	tagging	tagging	NOUN
fcis-6912	46	39	and	and	CCONJ
fcis-6912	46	40	named	name	VERB
fcis-6912	46	41	entity	entity	NOUN
fcis-6912	46	42	recognition	recognition	NOUN
fcis-6912	46	43	.	.	PUNCT
fcis-6912	47	1	sahu[14	sahu[14	PROPN
fcis-6912	47	2	]	]	PUNCT
fcis-6912	47	3	et	et	PROPN
fcis-6912	47	4	al	al	PROPN
fcis-6912	47	5	.	.	PROPN
fcis-6912	47	6	generated	generate	VERB
fcis-6912	47	7	word	word	NOUN
fcis-6912	47	8	embedding	embed	VERB
fcis-6912	47	9	features	feature	NOUN
fcis-6912	47	10	by	by	ADP
fcis-6912	47	11	cascading	cascade	VERB
fcis-6912	47	12	cnn	cnn	PROPN
fcis-6912	47	13	and	and	CCONJ
fcis-6912	47	14	rnn	rnn	PROPN
fcis-6912	47	15	,	,	PUNCT
fcis-6912	47	16	which	which	PRON
fcis-6912	47	17	can	can	AUX
fcis-6912	47	18	effectively	effectively	ADV
fcis-6912	47	19	reduce	reduce	VERB
fcis-6912	47	20	workload	workload	NOUN
fcis-6912	47	21	without	without	ADP
fcis-6912	47	22	too	too	ADV
fcis-6912	47	23	much	much	ADJ
fcis-6912	47	24	feature	feature	NOUN
fcis-6912	47	25	engineering	engineering	NOUN
fcis-6912	47	26	.	.	PUNCT
fcis-6912	48	1	vaswani	vaswani	NOUN
fcis-6912	49	1	[	[	X
fcis-6912	49	2	15	15	NUM
fcis-6912	49	3	]	]	X
fcis-6912	49	4	et	et	PROPN
fcis-6912	49	5	al	al	PROPN
fcis-6912	49	6	.	.	PROPN
fcis-6912	49	7	proposed	propose	VERB
fcis-6912	49	8	the	the	DET
fcis-6912	49	9	transformer	transformer	NOUN
fcis-6912	49	10	model	model	NOUN
fcis-6912	49	11	,	,	PUNCT
fcis-6912	49	12	which	which	PRON
fcis-6912	49	13	abandons	abandon	VERB
fcis-6912	49	14	traditional	traditional	ADJ
fcis-6912	49	15	cnn	cnn	PROPN
fcis-6912	49	16	and	and	CCONJ
fcis-6912	49	17	rnn	rnn	PROPN
fcis-6912	49	18	,	,	PUNCT
fcis-6912	49	19	and	and	CCONJ
fcis-6912	49	20	the	the	DET
fcis-6912	49	21	entire	entire	ADJ
fcis-6912	49	22	network	network	NOUN
fcis-6912	49	23	structure	structure	NOUN
fcis-6912	49	24	is	be	AUX
fcis-6912	49	25	composed	compose	VERB
fcis-6912	49	26	entirely	entirely	ADV
fcis-6912	49	27	of	of	ADP
fcis-6912	49	28	the	the	DET
fcis-6912	49	29	attention	attention	NOUN
fcis-6912	49	30	mechanism	mechanism	NOUN
fcis-6912	49	31	,	,	PUNCT
fcis-6912	49	32	which	which	PRON
fcis-6912	49	33	reduces	reduce	VERB
fcis-6912	49	34	complexity	complexity	NOUN
fcis-6912	49	35	and	and	CCONJ
fcis-6912	49	36	allows	allow	VERB
fcis-6912	49	37	for	for	ADP
fcis-6912	49	38	parallel	parallel	ADJ
fcis-6912	49	39	computation	computation	NOUN
fcis-6912	49	40	.	.	PUNCT
fcis-6912	50	1	yan	yan	PROPN
fcis-6912	51	1	[	[	X
fcis-6912	51	2	16	16	NUM
fcis-6912	51	3	]	]	PUNCT
fcis-6912	51	4	et	et	PROPN
fcis-6912	51	5	al	al	PROPN
fcis-6912	51	6	.	.	PROPN
fcis-6912	51	7	optimized	optimize	VERB
fcis-6912	51	8	the	the	DET
fcis-6912	51	9	transformer	transformer	NOUN
fcis-6912	51	10	-	-	PUNCT
fcis-6912	51	11	based	base	VERB
fcis-6912	51	12	model	model	NOUN
fcis-6912	51	13	and	and	CCONJ
fcis-6912	51	14	proposed	propose	VERB
fcis-6912	51	15	the	the	DET
fcis-6912	51	16	tener	tener	PROPN
fcis-6912	51	17	model	model	PROPN
fcis-6912	51	18	.	.	PUNCT
fcis-6912	52	1	incorporating	incorporate	VERB
fcis-6912	52	2	language	language	NOUN
fcis-6912	52	3	features	feature	NOUN
fcis-6912	52	4	such	such	ADJ
fcis-6912	52	5	as	as	ADP
fcis-6912	52	6	phonetic	phonetic	ADJ
fcis-6912	52	7	and	and	CCONJ
fcis-6912	52	8	character	character	NOUN
fcis-6912	52	9	features	feature	NOUN
fcis-6912	52	10	into	into	ADP
fcis-6912	52	11	the	the	DET
fcis-6912	52	12	model	model	NOUN
fcis-6912	52	13	is	be	AUX
fcis-6912	52	14	also	also	ADV
fcis-6912	52	15	a	a	DET
fcis-6912	52	16	way	way	NOUN
fcis-6912	52	17	to	to	PART
fcis-6912	52	18	improve	improve	VERB
fcis-6912	52	19	the	the	DET
fcis-6912	52	20	model	model	NOUN
fcis-6912	52	21	.	.	PUNCT
fcis-6912	53	1	bharadwaj	bharadwaj	PROPN
fcis-6912	54	1	[	[	X
fcis-6912	54	2	17	17	NUM
fcis-6912	54	3	]	]	PUNCT
fcis-6912	54	4	et	et	PROPN
fcis-6912	54	5	al	al	PROPN
fcis-6912	54	6	.	.	PROPN
fcis-6912	54	7	used	use	VERB
fcis-6912	54	8	lstm	lstm	NOUN
fcis-6912	54	9	to	to	PART
fcis-6912	54	10	incorporate	incorporate	VERB
fcis-6912	54	11	phonetic	phonetic	ADJ
fcis-6912	54	12	features	feature	NOUN
fcis-6912	54	13	to	to	PART
fcis-6912	54	14	improve	improve	VERB
fcis-6912	54	15	the	the	DET
fcis-6912	54	16	model	model	NOUN
fcis-6912	54	17	's	's	PART
fcis-6912	54	18	effectiveness	effectiveness	NOUN
fcis-6912	54	19	.	.	PUNCT
fcis-6912	55	1	with	with	ADP
fcis-6912	55	2	the	the	DET
fcis-6912	55	3	development	development	NOUN
fcis-6912	55	4	of	of	ADP
fcis-6912	55	5	the	the	DET
fcis-6912	55	6	machine	machine	NOUN
fcis-6912	55	7	learning	learn	VERB
fcis-6912	55	8	field	field	NOUN
fcis-6912	55	9	,	,	PUNCT
fcis-6912	55	10	neural	neural	ADJ
fcis-6912	55	11	network	network	NOUN
fcis-6912	55	12	design	design	NOUN
fcis-6912	55	13	has	have	AUX
fcis-6912	55	14	become	become	VERB
fcis-6912	55	15	increasingly	increasingly	ADV
fcis-6912	55	16	complex	complex	ADJ
fcis-6912	55	17	,	,	PUNCT
fcis-6912	55	18	leading	lead	VERB
fcis-6912	55	19	to	to	ADP
fcis-6912	55	20	more	more	ADJ
fcis-6912	55	21	and	and	CCONJ
fcis-6912	55	22	more	more	ADJ
fcis-6912	55	23	workload	workload	NOUN
fcis-6912	55	24	.	.	PUNCT
fcis-6912	56	1	people	people	NOUN
fcis-6912	56	2	use	use	VERB
fcis-6912	56	3	a	a	DET
fcis-6912	56	4	large	large	ADJ
fcis-6912	56	5	amount	amount	NOUN
fcis-6912	56	6	of	of	ADP
fcis-6912	56	7	data	datum	NOUN
fcis-6912	56	8	to	to	PART
fcis-6912	56	9	train	train	VERB
fcis-6912	56	10	the	the	DET
fcis-6912	56	11	network	network	NOUN
fcis-6912	56	12	structure	structure	NOUN
fcis-6912	56	13	to	to	PART
fcis-6912	56	14	obtain	obtain	VERB
fcis-6912	56	15	pre	pre	ADJ
fcis-6912	56	16	-	-	ADJ
fcis-6912	56	17	trained	train	VERB
fcis-6912	56	18	models	model	NOUN
fcis-6912	56	19	,	,	PUNCT
fcis-6912	56	20	which	which	PRON
fcis-6912	56	21	can	can	AUX
fcis-6912	56	22	be	be	AUX
fcis-6912	56	23	fine	fine	ADV
fcis-6912	56	24	-	-	PUNCT
fcis-6912	56	25	tuned	tune	VERB
fcis-6912	56	26	on	on	ADP
fcis-6912	56	27	their	their	PRON
fcis-6912	56	28	own	own	ADJ
fcis-6912	56	29	dataset	dataset	NOUN
fcis-6912	56	30	,	,	PUNCT
fcis-6912	56	31	reducing	reduce	VERB
fcis-6912	56	32	workload	workload	NOUN
fcis-6912	56	33	.	.	PUNCT
fcis-6912	57	1	there	there	PRON
fcis-6912	57	2	are	be	VERB
fcis-6912	57	3	already	already	ADV
fcis-6912	57	4	many	many	ADJ
fcis-6912	57	5	pre	pre	ADJ
fcis-6912	57	6	-	-	ADJ
fcis-6912	57	7	trained	train	VERB
fcis-6912	57	8	models	model	NOUN
fcis-6912	57	9	,	,	PUNCT
fcis-6912	57	10	such	such	ADJ
fcis-6912	57	11	as	as	ADP
fcis-6912	57	12	the	the	DET
fcis-6912	57	13	bert	bert	NOUN
fcis-6912	57	14	[	[	X
fcis-6912	57	15	18	18	NUM
fcis-6912	57	16	]	]	PUNCT
fcis-6912	57	17	model	model	NOUN
fcis-6912	57	18	proposed	propose	VERB
fcis-6912	57	19	by	by	ADP
fcis-6912	57	20	devlin	devlin	PROPN
fcis-6912	57	21	et	et	PROPN
fcis-6912	57	22	al	al	PROPN
fcis-6912	57	23	.	.	PROPN
fcis-6912	57	24	based	base	VERB
fcis-6912	57	25	on	on	ADP
fcis-6912	57	26	transformer	transformer	NOUN
fcis-6912	57	27	,	,	PUNCT
fcis-6912	57	28	the	the	DET
fcis-6912	57	29	roberta	roberta	PROPN
fcis-6912	57	30	[	[	X
fcis-6912	57	31	19	19	NUM
fcis-6912	57	32	]	]	PUNCT
fcis-6912	57	33	model	model	NOUN
fcis-6912	57	34	proposed	propose	VERB
fcis-6912	57	35	by	by	ADP
fcis-6912	57	36	the	the	DET
fcis-6912	57	37	facebook	facebook	PROPN
fcis-6912	57	38	team	team	NOUN
fcis-6912	57	39	the	the	DET
fcis-6912	57	40	following	following	ADJ
fcis-6912	57	41	year	year	NOUN
fcis-6912	57	42	,	,	PUNCT
fcis-6912	57	43	and	and	CCONJ
fcis-6912	57	44	the	the	DET
fcis-6912	57	45	albert	albert	PROPN
fcis-6912	57	46	[	[	X
fcis-6912	57	47	20	20	NUM
fcis-6912	57	48	]	]	PUNCT
fcis-6912	57	49	model	model	NOUN
fcis-6912	57	50	proposed	propose	VERB
fcis-6912	57	51	by	by	ADP
fcis-6912	57	52	lan	lan	PROPN
fcis-6912	57	53	et	et	PROPN
fcis-6912	57	54	al	al	PROPN
fcis-6912	57	55	.	.	PROPN
fcis-6912	57	56	improvements	improvement	NOUN
fcis-6912	57	57	to	to	ADP
fcis-6912	57	58	the	the	DET
fcis-6912	57	59	bert	bert	PROPN
fcis-6912	57	60	model	model	NOUN
fcis-6912	57	61	are	be	AUX
fcis-6912	57	62	mainly	mainly	ADV
fcis-6912	57	63	achieved	achieve	VERB
fcis-6912	57	64	by	by	ADP
fcis-6912	57	65	modifying	modify	VERB
fcis-6912	57	66	the	the	DET
fcis-6912	57	67	next	next	ADJ
fcis-6912	57	68	sentence	sentence	NOUN
fcis-6912	57	69	prediction	prediction	NOUN
fcis-6912	57	70	and	and	CCONJ
fcis-6912	57	71	masked	mask	VERB
fcis-6912	57	72	lm	lm	INTJ
fcis-6912	57	73	to	to	PART
fcis-6912	57	74	improve	improve	VERB
fcis-6912	57	75	model	model	NOUN
fcis-6912	57	76	performance	performance	NOUN
fcis-6912	57	77	[	[	X
fcis-6912	57	78	21	21	NUM
fcis-6912	57	79	]	]	PUNCT
fcis-6912	57	80	.	.	PUNCT
fcis-6912	58	1	dong	dong	PROPN
fcis-6912	59	1	[	[	X
fcis-6912	59	2	22	22	NUM
fcis-6912	59	3	]	]	PUNCT
fcis-6912	59	4	et	et	PROPN
fcis-6912	59	5	al	al	PROPN
fcis-6912	59	6	.	.	PROPN
fcis-6912	59	7	proposed	propose	VERB
fcis-6912	59	8	the	the	DET
fcis-6912	59	9	unilm	unilm	NOUN
fcis-6912	59	10	model	model	NOUN
fcis-6912	59	11	,	,	PUNCT
fcis-6912	59	12	and	and	CCONJ
fcis-6912	59	13	song	song	NOUN
fcis-6912	59	14	[	[	X
fcis-6912	59	15	23	23	NUM
fcis-6912	59	16	]	]	PUNCT
fcis-6912	59	17	et	et	PROPN
fcis-6912	59	18	al	al	PROPN
fcis-6912	59	19	.	.	PROPN
fcis-6912	59	20	proposed	propose	VERB
fcis-6912	59	21	the	the	DET
fcis-6912	59	22	mass	mass	NOUN
fcis-6912	59	23	model	model	NOUN
fcis-6912	59	24	.	.	PUNCT
fcis-6912	60	1	tsai	tsai	PROPN
fcis-6912	60	2	[	[	X
fcis-6912	60	3	24	24	NUM
fcis-6912	60	4	]	]	PUNCT
fcis-6912	60	5	et	et	PROPN
fcis-6912	60	6	al	al	PROPN
fcis-6912	60	7	.	.	PROPN
fcis-6912	60	8	proposed	propose	VERB
fcis-6912	60	9	a	a	DET
fcis-6912	60	10	bertbased	bertbase	VERB
fcis-6912	60	11	model	model	NOUN
fcis-6912	60	12	and	and	CCONJ
fcis-6912	60	13	used	use	VERB
fcis-6912	60	14	knowledge	knowledge	NOUN
fcis-6912	60	15	distillation	distillation	NOUN
fcis-6912	60	16	to	to	PART
fcis-6912	60	17	improve	improve	VERB
fcis-6912	60	18	model	model	NOUN
fcis-6912	60	19	training	training	NOUN
fcis-6912	60	20	time	time	NOUN
fcis-6912	60	21	.	.	PUNCT
fcis-6912	61	1	nested	nest	VERB
fcis-6912	61	2	named	name	VERB
fcis-6912	61	3	entities	entity	NOUN
fcis-6912	61	4	refer	refer	VERB
fcis-6912	61	5	to	to	ADP
fcis-6912	61	6	the	the	DET
fcis-6912	61	7	phenomenon	phenomenon	NOUN
fcis-6912	61	8	of	of	ADP
fcis-6912	61	9	entities	entity	NOUN
fcis-6912	61	10	nested	nest	VERB
fcis-6912	61	11	in	in	ADP
fcis-6912	61	12	entities	entity	NOUN
fcis-6912	61	13	.	.	PUNCT
fcis-6912	62	1	there	there	PRON
fcis-6912	62	2	are	be	VERB
fcis-6912	62	3	more	more	ADJ
fcis-6912	62	4	studies	study	NOUN
fcis-6912	62	5	on	on	ADP
fcis-6912	62	6	this	this	DET
fcis-6912	62	7	aspect	aspect	NOUN
fcis-6912	62	8	abroad	abroad	ADV
fcis-6912	62	9	,	,	PUNCT
fcis-6912	62	10	and	and	CCONJ
fcis-6912	62	11	the	the	DET
fcis-6912	62	12	methods	method	NOUN
fcis-6912	62	13	for	for	ADP
fcis-6912	62	14	nested	nested	ADJ
fcis-6912	62	15	named	name	VERB
fcis-6912	62	16	entities	entity	NOUN
fcis-6912	62	17	can	can	AUX
fcis-6912	62	18	be	be	AUX
fcis-6912	62	19	divided	divide	VERB
fcis-6912	62	20	into	into	ADP
fcis-6912	62	21	several	several	ADJ
fcis-6912	62	22	types	type	NOUN
fcis-6912	62	23	:	:	PUNCT
fcis-6912	62	24	based	base	VERB
fcis-6912	62	25	on	on	ADP
fcis-6912	62	26	hypergraphs	hypergraph	NOUN
fcis-6912	62	27	,	,	PUNCT
fcis-6912	62	28	stack	stack	NOUN
fcis-6912	62	29	-	-	PUNCT
fcis-6912	62	30	based	base	VERB
fcis-6912	62	31	model	model	NOUN
fcis-6912	62	32	region	region	NOUN
fcis-6912	62	33	models	model	NOUN
fcis-6912	62	34	,	,	PUNCT
fcis-6912	62	35	and	and	CCONJ
fcis-6912	62	36	based	base	VERB
fcis-6912	62	37	on	on	ADP
fcis-6912	62	38	reading	read	VERB
fcis-6912	62	39	comprehension	comprehension	NOUN
fcis-6912	62	40	.	.	PUNCT
fcis-6912	63	1	lu	lu	PROPN
fcis-6912	64	1	[	[	X
fcis-6912	64	2	25	25	NUM
fcis-6912	64	3	]	]	PUNCT
fcis-6912	64	4	et	et	PROPN
fcis-6912	64	5	al	al	PROPN
fcis-6912	64	6	.	.	PROPN
fcis-6912	65	1	first	first	PROPN
fcis-6912	65	2	proposed	propose	VERB
fcis-6912	65	3	a	a	DET
fcis-6912	65	4	method	method	NOUN
fcis-6912	65	5	for	for	ADP
fcis-6912	65	6	nested	nested	ADJ
fcis-6912	65	7	named	name	VERB
fcis-6912	65	8	entity	entity	NOUN
fcis-6912	65	9	recognition	recognition	NOUN
fcis-6912	65	10	based	base	VERB
fcis-6912	65	11	on	on	ADP
fcis-6912	65	12	hypergraphs	hypergraph	NOUN
fcis-6912	65	13	in	in	ADP
fcis-6912	65	14	2015	2015	NUM
fcis-6912	65	15	,	,	PUNCT
fcis-6912	65	16	which	which	PRON
fcis-6912	65	17	merges	merge	VERB
fcis-6912	65	18	tokens	token	NOUN
fcis-6912	65	19	with	with	ADP
fcis-6912	65	20	the	the	DET
fcis-6912	65	21	same	same	ADJ
fcis-6912	65	22	value	value	NOUN
fcis-6912	65	23	and	and	CCONJ
fcis-6912	65	24	makes	make	VERB
fcis-6912	65	25	each	each	DET
fcis-6912	65	26	token	token	VERB
fcis-6912	65	27	contain	contain	VERB
fcis-6912	65	28	an	an	DET
fcis-6912	65	29	o	o	NOUN
fcis-6912	65	30	label	label	NOUN
fcis-6912	65	31	,	,	PUNCT
fcis-6912	65	32	and	and	CCONJ
fcis-6912	65	33	then	then	ADV
fcis-6912	65	34	the	the	DET
fcis-6912	65	35	decoder	decoder	NOUN
fcis-6912	65	36	outputs	output	VERB
fcis-6912	65	37	all	all	DET
fcis-6912	65	38	possible	possible	ADJ
fcis-6912	65	39	labels	label	NOUN
fcis-6912	65	40	.	.	PUNCT
fcis-6912	66	1	muis[26	muis[26	X
fcis-6912	66	2	]	]	PUNCT
fcis-6912	66	3	et	et	PROPN
fcis-6912	66	4	al	al	PROPN
fcis-6912	66	5	.	.	PROPN
fcis-6912	66	6	proposed	propose	VERB
fcis-6912	66	7	a	a	DET
fcis-6912	66	8	multi	multi	ADJ
fcis-6912	66	9	-	-	ADJ
fcis-6912	66	10	graph	graph	NOUN
fcis-6912	66	11	representation	representation	NOUN
fcis-6912	66	12	method	method	NOUN
fcis-6912	66	13	,	,	PUNCT
fcis-6912	66	14	which	which	PRON
fcis-6912	66	15	uses	use	VERB
fcis-6912	66	16	separators	separator	NOUN
fcis-6912	66	17	to	to	PART
fcis-6912	66	18	detect	detect	VERB
fcis-6912	66	19	nested	nested	ADJ
fcis-6912	66	20	entities	entity	NOUN
fcis-6912	66	21	.	.	PUNCT
fcis-6912	67	1	ju	ju	NOUN
fcis-6912	68	1	[	[	X
fcis-6912	68	2	27	27	NUM
fcis-6912	68	3	]	]	PUNCT
fcis-6912	68	4	et	et	PROPN
fcis-6912	68	5	al	al	PROPN
fcis-6912	68	6	.	.	PROPN
fcis-6912	68	7	extracted	extract	VERB
fcis-6912	68	8	nested	nested	ADJ
fcis-6912	68	9	entities	entity	NOUN
fcis-6912	68	10	by	by	ADP
fcis-6912	68	11	stacking	stack	VERB
fcis-6912	68	12	bilstm+crf	bilstm+crf	NOUN
fcis-6912	68	13	layers	layer	NOUN
fcis-6912	68	14	.	.	PUNCT
fcis-6912	69	1	jue[28	jue[28	PROPN
fcis-6912	69	2	]	]	PUNCT
fcis-6912	70	1	et	et	PROPN
fcis-6912	70	2	al	al	PROPN
fcis-6912	70	3	.	.	PROPN
fcis-6912	70	4	stacked	stack	VERB
fcis-6912	70	5	each	each	PRON
fcis-6912	70	6	token	token	VERB
fcis-6912	70	7	from	from	ADP
fcis-6912	70	8	bottom	bottom	NOUN
fcis-6912	70	9	to	to	ADP
fcis-6912	70	10	top	top	NOUN
fcis-6912	70	11	,	,	PUNCT
fcis-6912	70	12	generating	generate	VERB
fcis-6912	70	13	longer	long	ADJ
fcis-6912	70	14	entities	entity	NOUN
fcis-6912	70	15	that	that	PRON
fcis-6912	70	16	exceed	exceed	VERB
fcis-6912	70	17	the	the	DET
fcis-6912	70	18	length	length	NOUN
fcis-6912	70	19	of	of	ADP
fcis-6912	70	20	the	the	DET
fcis-6912	70	21	lower	low	ADJ
fcis-6912	70	22	layers	layer	NOUN
fcis-6912	70	23	,	,	PUNCT
fcis-6912	70	24	and	and	CCONJ
fcis-6912	70	25	detected	detect	VERB
fcis-6912	70	26	all	all	DET
fcis-6912	70	27	entities	entity	NOUN
fcis-6912	70	28	in	in	ADP
fcis-6912	70	29	the	the	DET
fcis-6912	70	30	text	text	NOUN
fcis-6912	70	31	by	by	ADP
fcis-6912	70	32	traversing	traverse	VERB
fcis-6912	70	33	all	all	DET
fcis-6912	70	34	entities	entity	NOUN
fcis-6912	70	35	.	.	PUNCT
fcis-6912	71	1	named	name	VERB
fcis-6912	71	2	entity	entity	NOUN
fcis-6912	71	3	recognition	recognition	NOUN
fcis-6912	71	4	technology	technology	NOUN
fcis-6912	71	5	has	have	AUX
fcis-6912	71	6	gradually	gradually	ADV
fcis-6912	71	7	matured	mature	VERB
fcis-6912	71	8	,	,	PUNCT
fcis-6912	71	9	especially	especially	ADV
fcis-6912	71	10	with	with	ADP
fcis-6912	71	11	the	the	DET
fcis-6912	71	12	application	application	NOUN
fcis-6912	71	13	of	of	ADP
fcis-6912	71	14	neural	neural	ADJ
fcis-6912	71	15	networks	network	NOUN
fcis-6912	71	16	,	,	PUNCT
fcis-6912	71	17	which	which	PRON
fcis-6912	71	18	has	have	AUX
fcis-6912	71	19	led	lead	VERB
fcis-6912	71	20	to	to	ADP
fcis-6912	71	21	a	a	DET
fcis-6912	71	22	gradual	gradual	ADJ
fcis-6912	71	23	improvement	improvement	NOUN
fcis-6912	71	24	in	in	ADP
fcis-6912	71	25	the	the	DET
fcis-6912	71	26	f1	f1	ADJ
fcis-6912	71	27	score	score	NOUN
fcis-6912	71	28	of	of	ADP
fcis-6912	71	29	named	name	VERB
fcis-6912	71	30	entity	entity	NOUN
fcis-6912	71	31	recognition	recognition	NOUN
fcis-6912	71	32	.	.	PUNCT
fcis-6912	72	1	pretrained	pretraine	VERB
fcis-6912	72	2	models	model	NOUN
fcis-6912	72	3	can	can	AUX
fcis-6912	72	4	solve	solve	VERB
fcis-6912	72	5	certain	certain	ADJ
fcis-6912	72	6	problems	problem	NOUN
fcis-6912	72	7	,	,	PUNCT
fcis-6912	72	8	but	but	CCONJ
fcis-6912	72	9	new	new	ADJ
fcis-6912	72	10	issues	issue	NOUN
fcis-6912	72	11	have	have	AUX
fcis-6912	72	12	also	also	ADV
fcis-6912	72	13	emerged	emerge	VERB
fcis-6912	72	14	,	,	PUNCT
fcis-6912	72	15	such	such	ADJ
fcis-6912	72	16	as	as	ADP
fcis-6912	72	17	excessive	excessive	ADJ
fcis-6912	72	18	resource	resource	NOUN
fcis-6912	72	19	consumption	consumption	NOUN
fcis-6912	72	20	[	[	X
fcis-6912	72	21	29	29	NUM
fcis-6912	72	22	]	]	PUNCT
fcis-6912	72	23	.	.	PUNCT
fcis-6912	73	1	3	3	X
fcis-6912	73	2	.	.	X
fcis-6912	73	3	ernie	ernie	NOUN
fcis-6912	73	4	-	-	PUNCT
fcis-6912	73	5	bigru	bigru	PROPN
fcis-6912	73	6	-	-	PUNCT
fcis-6912	73	7	crf	crf	NOUN
fcis-6912	73	8	model	model	NOUN
fcis-6912	73	9	the	the	DET
fcis-6912	73	10	overall	overall	ADJ
fcis-6912	73	11	architecture	architecture	NOUN
fcis-6912	73	12	of	of	ADP
fcis-6912	73	13	the	the	DET
fcis-6912	73	14	ernie	ernie	NOUN
fcis-6912	73	15	-	-	PUNCT
fcis-6912	73	16	bigru	bigru	PROPN
fcis-6912	73	17	-	-	PUNCT
fcis-6912	73	18	crf	crf	NOUN
fcis-6912	73	19	model	model	NOUN
fcis-6912	73	20	is	be	AUX
fcis-6912	73	21	shown	show	VERB
fcis-6912	73	22	in	in	ADP
fcis-6912	73	23	figure	figure	NOUN
fcis-6912	73	24	1	1	NUM
fcis-6912	73	25	.	.	PUNCT
fcis-6912	74	1	the	the	DET
fcis-6912	74	2	model	model	NOUN
fcis-6912	74	3	first	first	ADV
fcis-6912	74	4	obtains	obtain	VERB
fcis-6912	74	5	the	the	DET
fcis-6912	74	6	semantic	semantic	ADJ
fcis-6912	74	7	representation	representation	NOUN
fcis-6912	74	8	of	of	ADP
fcis-6912	74	9	the	the	DET
fcis-6912	74	10	input	input	NOUN
fcis-6912	74	11	through	through	ADP
fcis-6912	74	12	the	the	DET
fcis-6912	74	13	ernie	ernie	NOUN
fcis-6912	74	14	pre	pre	VERB
fcis-6912	74	15	-	-	ADJ
fcis-6912	74	16	trained	train	VERB
fcis-6912	74	17	language	language	NOUN
fcis-6912	74	18	model	model	NOUN
fcis-6912	74	19	,	,	PUNCT
fcis-6912	74	20	which	which	PRON
fcis-6912	74	21	enhances	enhance	VERB
fcis-6912	74	22	knowledge	knowledge	NOUN
fcis-6912	74	23	-	-	PUNCT
fcis-6912	74	24	based	base	VERB
fcis-6912	74	25	semantic	semantic	ADJ
fcis-6912	74	26	representations	representation	NOUN
fcis-6912	74	27	.	.	PUNCT
fcis-6912	75	1	the	the	DET
fcis-6912	75	2	obtained	obtain	VERB
fcis-6912	75	3	word	word	NOUN
fcis-6912	75	4	vectors	vector	NOUN
fcis-6912	75	5	are	be	AUX
fcis-6912	75	6	inputted	inputte	VERB
fcis-6912	75	7	into	into	ADP
fcis-6912	75	8	a	a	DET
fcis-6912	75	9	bidirectional	bidirectional	ADJ
fcis-6912	75	10	gru	gru	NOUN
fcis-6912	75	11	layer	layer	NOUN
fcis-6912	75	12	to	to	PART
fcis-6912	75	13	extract	extract	VERB
fcis-6912	75	14	sentence	sentence	NOUN
fcis-6912	75	15	-	-	PUNCT
fcis-6912	75	16	level	level	NOUN
fcis-6912	75	17	features	feature	NOUN
fcis-6912	75	18	,	,	PUNCT
fcis-6912	75	19	and	and	CCONJ
fcis-6912	75	20	finally	finally	ADV
fcis-6912	75	21	,	,	PUNCT
fcis-6912	75	22	the	the	DET
fcis-6912	75	23	crf	crf	NOUN
fcis-6912	75	24	layer	layer	NOUN
fcis-6912	75	25	performs	perform	VERB
fcis-6912	75	26	sequence	sequence	NOUN
fcis-6912	75	27	labeling	labeling	NOUN
fcis-6912	75	28	to	to	PART
fcis-6912	75	29	obtain	obtain	VERB
fcis-6912	75	30	the	the	DET
fcis-6912	75	31	globally	globally	ADV
fcis-6912	75	32	optimal	optimal	ADJ
fcis-6912	75	33	label	label	NOUN
fcis-6912	75	34	sequence	sequence	NOUN
fcis-6912	75	35	.	.	PUNCT
fcis-6912	76	1	compared	compare	VERB
fcis-6912	76	2	with	with	ADP
fcis-6912	76	3	previous	previous	ADJ
fcis-6912	76	4	mainstream	mainstream	NOUN
fcis-6912	76	5	named	name	VERB
fcis-6912	76	6	entity	entity	NOUN
fcis-6912	76	7	recognition	recognition	NOUN
fcis-6912	76	8	models	model	NOUN
fcis-6912	76	9	,	,	PUNCT
fcis-6912	76	10	the	the	DET
fcis-6912	76	11	most	most	ADV
fcis-6912	76	12	significant	significant	ADJ
fcis-6912	76	13	difference	difference	NOUN
fcis-6912	76	14	of	of	ADP
fcis-6912	76	15	the	the	DET
fcis-6912	76	16	ernie	ernie	NOUN
fcis-6912	76	17	-	-	PUNCT
fcis-6912	76	18	bigru	bigru	PROPN
fcis-6912	76	19	-	-	PUNCT
fcis-6912	76	20	crf	crf	NOUN
fcis-6912	76	21	model	model	NOUN
fcis-6912	76	22	is	be	AUX
fcis-6912	76	23	the	the	DET
fcis-6912	76	24	incorporation	incorporation	NOUN
fcis-6912	76	25	of	of	ADP
fcis-6912	76	26	knowledge	knowledge	NOUN
fcis-6912	76	27	-	-	PUNCT
fcis-6912	76	28	enhanced	enhance	VERB
fcis-6912	76	29	semantic	semantic	ADJ
fcis-6912	76	30	representations	representation	NOUN
fcis-6912	76	31	from	from	ADP
fcis-6912	76	32	the	the	DET
fcis-6912	76	33	ernie	ernie	NOUN
fcis-6912	76	34	pre	pre	VERB
fcis-6912	76	35	-	-	ADJ
fcis-6912	76	36	trained	train	VERB
fcis-6912	76	37	language	language	NOUN
fcis-6912	76	38	model	model	NOUN
fcis-6912	76	39	.	.	PUNCT
fcis-6912	77	1	the	the	DET
fcis-6912	77	2	ernie	ernie	PROPN
fcis-6912	77	3	model	model	PROPN
fcis-6912	77	4	learns	learn	VERB
fcis-6912	77	5	a	a	DET
fcis-6912	77	6	comprehensive	comprehensive	ADJ
fcis-6912	77	7	semantic	semantic	ADJ
fcis-6912	77	8	representation	representation	NOUN
fcis-6912	77	9	of	of	ADP
fcis-6912	77	10	concepts	concept	NOUN
fcis-6912	77	11	by	by	ADP
fcis-6912	77	12	masking	mask	VERB
fcis-6912	77	13	semantic	semantic	ADJ
fcis-6912	77	14	units	unit	NOUN
fcis-6912	77	15	such	such	ADJ
fcis-6912	77	16	as	as	ADP
fcis-6912	77	17	words	word	NOUN
fcis-6912	77	18	and	and	CCONJ
fcis-6912	77	19	entities	entity	NOUN
fcis-6912	77	20	.	.	PUNCT
fcis-6912	78	1	this	this	DET
fcis-6912	78	2	representation	representation	NOUN
fcis-6912	78	3	captures	capture	VERB
fcis-6912	78	4	the	the	DET
fcis-6912	78	5	ambiguity	ambiguity	NOUN
fcis-6912	78	6	of	of	ADP
fcis-6912	78	7	words	word	NOUN
fcis-6912	78	8	and	and	CCONJ
fcis-6912	78	9	enhances	enhance	VERB
fcis-6912	78	10	the	the	DET
fcis-6912	78	11	semantic	semantic	ADJ
fcis-6912	78	12	representation	representation	NOUN
fcis-6912	78	13	capacity	capacity	NOUN
fcis-6912	78	14	of	of	ADP
fcis-6912	78	15	the	the	DET
fcis-6912	78	16	model	model	NOUN
fcis-6912	78	17	.	.	PUNCT
fcis-6912	79	1	fig	fig	NOUN
fcis-6912	79	2	.	.	PUNCT
fcis-6912	80	1	1	1	NUM
fcis-6912	80	2	named	name	VERB
fcis-6912	80	3	entity	entity	NOUN
fcis-6912	80	4	recognition	recognition	NOUN
fcis-6912	80	5	framework	framework	NOUN
fcis-6912	80	6	based	base	VERB
fcis-6912	80	7	on	on	ADP
fcis-6912	80	8	ernie	ernie	PROPN
fcis-6912	80	9	bigru	bigru	PROPN
fcis-6912	80	10	-	-	PUNCT
fcis-6912	80	11	crf	crf	PROPN
fcis-6912	80	12	3.1	3.1	NUM
fcis-6912	80	13	.	.	PUNCT
fcis-6912	81	1	ernie	ernie	PROPN
fcis-6912	81	2	pre	pre	VERB
fcis-6912	81	3	-	-	ADJ
fcis-6912	81	4	trained	train	VERB
fcis-6912	81	5	language	language	NOUN
fcis-6912	81	6	model	model	NOUN
fcis-6912	81	7	ernie	ernie	PROPN
fcis-6912	81	8	is	be	AUX
fcis-6912	81	9	a	a	DET
fcis-6912	81	10	knowledge	knowledge	NOUN
fcis-6912	81	11	-	-	PUNCT
fcis-6912	81	12	enhanced	enhance	VERB
fcis-6912	81	13	semantic	semantic	ADJ
fcis-6912	81	14	representation	representation	NOUN
fcis-6912	81	15	model	model	NOUN
fcis-6912	81	16	that	that	PRON
fcis-6912	81	17	models	model	NOUN
fcis-6912	81	18	prior	prior	ADV
fcis-6912	81	19	semantic	semantic	ADJ
fcis-6912	81	20	knowledge	knowledge	NOUN
fcis-6912	81	21	of	of	ADP
fcis-6912	81	22	words	word	NOUN
fcis-6912	81	23	,	,	PUNCT
fcis-6912	81	24	entities	entity	NOUN
fcis-6912	81	25	,	,	PUNCT
fcis-6912	81	26	and	and	CCONJ
fcis-6912	81	27	entity	entity	NOUN
fcis-6912	81	28	relationships	relationship	NOUN
fcis-6912	81	29	in	in	ADP
fcis-6912	81	30	massive	massive	ADJ
fcis-6912	81	31	data	datum	NOUN
fcis-6912	81	32	to	to	PART
fcis-6912	81	33	learn	learn	VERB
fcis-6912	81	34	a	a	DET
fcis-6912	81	35	comprehensive	comprehensive	ADJ
fcis-6912	81	36	semantic	semantic	ADJ
fcis-6912	81	37	representation	representation	NOUN
fcis-6912	81	38	of	of	ADP
fcis-6912	81	39	concepts	concept	NOUN
fcis-6912	81	40	[	[	X
fcis-6912	81	41	30	30	NUM
fcis-6912	81	42	]	]	PUNCT
fcis-6912	81	43	.	.	PUNCT
fcis-6912	82	1	ernie	ernie	PROPN
fcis-6912	82	2	and	and	CCONJ
fcis-6912	82	3	bert	bert	PROPN
fcis-6912	82	4	are	be	AUX
fcis-6912	82	5	both	both	PRON
fcis-6912	82	6	pre	pre	ADJ
fcis-6912	82	7	-	-	ADJ
fcis-6912	82	8	trained	train	VERB
fcis-6912	82	9	language	language	NOUN
fcis-6912	82	10	models	model	NOUN
fcis-6912	82	11	constructed	construct	VERB
fcis-6912	82	12	with	with	ADP
fcis-6912	82	13	a	a	DET
fcis-6912	82	14	multi	multi	ADJ
fcis-6912	82	15	-	-	ADJ
fcis-6912	82	16	layer	layer	ADJ
fcis-6912	82	17	bidirectional	bidirectional	ADJ
fcis-6912	82	18	transformer	transformer	NOUN
fcis-6912	82	19	encoder	encoder	NOUN
fcis-6912	82	20	as	as	ADP
fcis-6912	82	21	the	the	DET
fcis-6912	82	22	basic	basic	ADJ
fcis-6912	82	23	unit	unit	NOUN
fcis-6912	82	24	,	,	PUNCT
fcis-6912	82	25	as	as	SCONJ
fcis-6912	82	26	shown	show	VERB
fcis-6912	82	27	in	in	ADP
fcis-6912	82	28	figure	figure	NOUN
fcis-6912	82	29	2	2	NUM
fcis-6912	82	30	[	[	X
fcis-6912	82	31	31	31	NUM
fcis-6912	82	32	]	]	PUNCT
fcis-6912	82	33	.	.	PUNCT
fcis-6912	83	1	fig	fig	NOUN
fcis-6912	83	2	.	.	PUNCT
fcis-6912	84	1	2	2	NUM
fcis-6912	84	2	structure	structure	NOUN
fcis-6912	84	3	of	of	ADP
fcis-6912	84	4	ernie	ernie	PROPN
fcis-6912	84	5	23	23	NUM
fcis-6912	84	6	ernie	ernie	PROPN
fcis-6912	84	7	consists	consist	VERB
fcis-6912	84	8	of	of	ADP
fcis-6912	84	9	two	two	NUM
fcis-6912	84	10	parts	part	NOUN
fcis-6912	84	11	:	:	PUNCT
fcis-6912	84	12	encoding	encoding	NOUN
fcis-6912	84	13	and	and	CCONJ
fcis-6912	84	14	knowledge	knowledge	NOUN
fcis-6912	84	15	integration	integration	NOUN
fcis-6912	84	16	.	.	PUNCT
fcis-6912	85	1	in	in	ADP
fcis-6912	85	2	the	the	DET
fcis-6912	85	3	encoding	encoding	NOUN
fcis-6912	85	4	part	part	NOUN
fcis-6912	85	5	,	,	PUNCT
fcis-6912	85	6	the	the	DET
fcis-6912	85	7	transformer	transformer	NOUN
fcis-6912	85	8	encoder	encoder	NOUN
fcis-6912	85	9	is	be	AUX
fcis-6912	85	10	used	use	VERB
fcis-6912	85	11	to	to	PART
fcis-6912	85	12	generate	generate	VERB
fcis-6912	85	13	text	text	NOUN
fcis-6912	85	14	word	word	NOUN
fcis-6912	85	15	vectors	vector	NOUN
fcis-6912	85	16	that	that	PRON
fcis-6912	85	17	fuse	fuse	VERB
fcis-6912	85	18	contextual	contextual	ADJ
fcis-6912	85	19	semantic	semantic	ADJ
fcis-6912	85	20	information	information	NOUN
fcis-6912	85	21	.	.	PUNCT
fcis-6912	86	1	as	as	SCONJ
fcis-6912	86	2	shown	show	VERB
fcis-6912	86	3	in	in	ADP
fcis-6912	86	4	the	the	DET
fcis-6912	86	5	diagram	diagram	NOUN
fcis-6912	86	6	,	,	PUNCT
fcis-6912	86	7	the	the	DET
fcis-6912	86	8	transformer	transformer	NOUN
fcis-6912	86	9	encoder	encoder	NOUN
fcis-6912	86	10	is	be	AUX
fcis-6912	86	11	modeled	model	VERB
fcis-6912	86	12	based	base	VERB
fcis-6912	86	13	on	on	ADP
fcis-6912	86	14	an	an	DET
fcis-6912	86	15	attention	attention	NOUN
fcis-6912	86	16	mechanism	mechanism	NOUN
fcis-6912	86	17	.	.	PUNCT
fcis-6912	87	1	the	the	DET
fcis-6912	87	2	attention	attention	NOUN
fcis-6912	87	3	mechanism	mechanism	NOUN
fcis-6912	87	4	can	can	AUX
fcis-6912	87	5	reflect	reflect	VERB
fcis-6912	87	6	the	the	DET
fcis-6912	87	7	importance	importance	NOUN
fcis-6912	87	8	of	of	ADP
fcis-6912	87	9	different	different	ADJ
fcis-6912	87	10	vocabulary	vocabulary	NOUN
fcis-6912	87	11	in	in	ADP
fcis-6912	87	12	the	the	DET
fcis-6912	87	13	text	text	NOUN
fcis-6912	87	14	and	and	CCONJ
fcis-6912	87	15	improve	improve	VERB
fcis-6912	87	16	the	the	DET
fcis-6912	87	17	weight	weight	NOUN
fcis-6912	87	18	of	of	ADP
fcis-6912	87	19	information	information	NOUN
fcis-6912	87	20	relevant	relevant	ADJ
fcis-6912	87	21	to	to	PART
fcis-6912	87	22	defect	defect	VERB
fcis-6912	87	23	text	text	NOUN
fcis-6912	87	24	classification	classification	NOUN
fcis-6912	87	25	,	,	PUNCT
fcis-6912	87	26	thereby	thereby	ADV
fcis-6912	87	27	further	far	ADV
fcis-6912	87	28	improving	improve	VERB
fcis-6912	87	29	the	the	DET
fcis-6912	87	30	accuracy	accuracy	NOUN
fcis-6912	87	31	of	of	ADP
fcis-6912	87	32	feature	feature	NOUN
fcis-6912	87	33	extraction	extraction	NOUN
fcis-6912	87	34	by	by	ADP
fcis-6912	87	35	the	the	DET
fcis-6912	87	36	model	model	NOUN
fcis-6912	87	37	.	.	PUNCT
fcis-6912	88	1	in	in	ADP
fcis-6912	88	2	the	the	DET
fcis-6912	88	3	knowledge	knowledge	NOUN
fcis-6912	88	4	integration	integration	NOUN
fcis-6912	88	5	part	part	NOUN
fcis-6912	88	6	,	,	PUNCT
fcis-6912	88	7	ernie	ernie	PROPN
fcis-6912	88	8	proposes	propose	VERB
fcis-6912	88	9	a	a	DET
fcis-6912	88	10	multi	multi	ADJ
fcis-6912	88	11	-	-	ADJ
fcis-6912	88	12	stage	stage	ADJ
fcis-6912	88	13	knowledge	knowledge	NOUN
fcis-6912	88	14	masking	masking	NOUN
fcis-6912	88	15	strategy	strategy	NOUN
fcis-6912	88	16	that	that	PRON
fcis-6912	88	17	integrates	integrate	VERB
fcis-6912	88	18	semantic	semantic	ADJ
fcis-6912	88	19	knowledge	knowledge	NOUN
fcis-6912	88	20	of	of	ADP
fcis-6912	88	21	defect	defect	NOUN
fcis-6912	88	22	text	text	NOUN
fcis-6912	88	23	at	at	ADP
fcis-6912	88	24	the	the	DET
fcis-6912	88	25	character	character	NOUN
fcis-6912	88	26	,	,	PUNCT
fcis-6912	88	27	phrase	phrase	NOUN
fcis-6912	88	28	,	,	PUNCT
fcis-6912	88	29	and	and	CCONJ
fcis-6912	88	30	entity	entity	NOUN
fcis-6912	88	31	levels	level	NOUN
fcis-6912	88	32	.	.	PUNCT
fcis-6912	89	1	knowledge	knowledge	NOUN
fcis-6912	89	2	masking	masking	NOUN
fcis-6912	89	3	refers	refer	VERB
fcis-6912	89	4	to	to	PART
fcis-6912	89	5	randomly	randomly	ADV
fcis-6912	89	6	masking	mask	VERB
fcis-6912	89	7	some	some	DET
fcis-6912	89	8	characters	character	NOUN
fcis-6912	89	9	and	and	CCONJ
fcis-6912	89	10	training	train	VERB
fcis-6912	89	11	the	the	DET
fcis-6912	89	12	model	model	NOUN
fcis-6912	89	13	to	to	PART
fcis-6912	89	14	predict	predict	VERB
fcis-6912	89	15	the	the	DET
fcis-6912	89	16	masked	masked	ADJ
fcis-6912	89	17	part	part	NOUN
fcis-6912	89	18	,	,	PUNCT
fcis-6912	89	19	thereby	thereby	ADV
fcis-6912	89	20	effectively	effectively	ADV
fcis-6912	89	21	learning	learn	VERB
fcis-6912	89	22	the	the	DET
fcis-6912	89	23	contextual	contextual	ADJ
fcis-6912	89	24	information	information	NOUN
fcis-6912	89	25	of	of	ADP
fcis-6912	89	26	the	the	DET
fcis-6912	89	27	masked	masked	ADJ
fcis-6912	89	28	part	part	NOUN
fcis-6912	89	29	.	.	PUNCT
fcis-6912	90	1	compared	compare	VERB
fcis-6912	90	2	to	to	ADP
fcis-6912	90	3	the	the	DET
fcis-6912	90	4	single	single	ADJ
fcis-6912	90	5	-	-	PUNCT
fcis-6912	90	6	character	character	NOUN
fcis-6912	90	7	masking	masking	NOUN
fcis-6912	90	8	strategy	strategy	NOUN
fcis-6912	90	9	of	of	ADP
fcis-6912	90	10	the	the	DET
fcis-6912	90	11	bert	bert	PROPN
fcis-6912	90	12	model	model	PROPN
fcis-6912	90	13	,	,	PUNCT
fcis-6912	90	14	ernie	ernie	PROPN
fcis-6912	90	15	introduces	introduce	VERB
fcis-6912	90	16	three	three	NUM
fcis-6912	90	17	levels	level	NOUN
fcis-6912	90	18	of	of	ADP
fcis-6912	90	19	masking	masking	NOUN
fcis-6912	90	20	,	,	PUNCT
fcis-6912	90	21	including	include	VERB
fcis-6912	90	22	characters	character	NOUN
fcis-6912	90	23	,	,	PUNCT
fcis-6912	90	24	phrases	phrase	NOUN
fcis-6912	90	25	,	,	PUNCT
fcis-6912	90	26	and	and	CCONJ
fcis-6912	90	27	entities	entity	NOUN
fcis-6912	90	28	,	,	PUNCT
fcis-6912	90	29	and	and	CCONJ
fcis-6912	90	30	through	through	ADP
fcis-6912	90	31	a	a	DET
fcis-6912	90	32	multi	multi	ADJ
fcis-6912	90	33	-	-	ADJ
fcis-6912	90	34	stage	stage	ADJ
fcis-6912	90	35	knowledge	knowledge	NOUN
fcis-6912	90	36	masking	masking	NOUN
fcis-6912	90	37	strategy	strategy	NOUN
fcis-6912	90	38	,	,	PUNCT
fcis-6912	90	39	ernie	ernie	PROPN
fcis-6912	90	40	can	can	AUX
fcis-6912	90	41	generate	generate	VERB
fcis-6912	90	42	word	word	NOUN
fcis-6912	90	43	vectors	vector	NOUN
fcis-6912	90	44	that	that	PRON
fcis-6912	90	45	contain	contain	VERB
fcis-6912	90	46	rich	rich	ADJ
fcis-6912	90	47	semantic	semantic	ADJ
fcis-6912	90	48	information	information	NOUN
fcis-6912	90	49	of	of	ADP
fcis-6912	90	50	the	the	DET
fcis-6912	90	51	defect	defect	NOUN
fcis-6912	90	52	text	text	NOUN
fcis-6912	90	53	and	and	CCONJ
fcis-6912	90	54	effectively	effectively	ADV
fcis-6912	90	55	preserve	preserve	VERB
fcis-6912	90	56	the	the	DET
fcis-6912	90	57	correlation	correlation	NOUN
fcis-6912	90	58	between	between	ADP
fcis-6912	90	59	various	various	ADJ
fcis-6912	90	60	components	component	NOUN
fcis-6912	90	61	of	of	ADP
fcis-6912	90	62	the	the	DET
fcis-6912	90	63	defect	defect	NOUN
fcis-6912	90	64	text	text	NOUN
fcis-6912	90	65	,	,	PUNCT
fcis-6912	90	66	thus	thus	ADV
fcis-6912	90	67	ensuring	ensure	VERB
fcis-6912	90	68	that	that	SCONJ
fcis-6912	90	69	important	important	ADJ
fcis-6912	90	70	semantic	semantic	ADJ
fcis-6912	90	71	information	information	NOUN
fcis-6912	90	72	is	be	AUX
fcis-6912	90	73	not	not	PART
fcis-6912	90	74	lost	lose	VERB
fcis-6912	90	75	.	.	PUNCT
fcis-6912	91	1	3.2	3.2	NUM
fcis-6912	91	2	.	.	PUNCT
fcis-6912	92	1	recurrent	recurrent	ADJ
fcis-6912	92	2	neural	neural	ADJ
fcis-6912	92	3	network	network	NOUN
fcis-6912	92	4	traditional	traditional	ADJ
fcis-6912	92	5	machine	machine	NOUN
fcis-6912	92	6	learning	learning	NOUN
fcis-6912	92	7	methods	method	NOUN
fcis-6912	92	8	rely	rely	VERB
fcis-6912	92	9	on	on	ADP
fcis-6912	92	10	manual	manual	ADJ
fcis-6912	92	11	feature	feature	NOUN
fcis-6912	92	12	extraction	extraction	NOUN
fcis-6912	92	13	before	before	ADP
fcis-6912	92	14	text	text	NOUN
fcis-6912	92	15	classification	classification	NOUN
fcis-6912	92	16	,	,	PUNCT
fcis-6912	92	17	which	which	PRON
fcis-6912	92	18	can	can	AUX
fcis-6912	92	19	easily	easily	ADV
fcis-6912	92	20	lead	lead	VERB
fcis-6912	92	21	to	to	ADP
fcis-6912	92	22	loss	loss	NOUN
fcis-6912	92	23	of	of	ADP
fcis-6912	92	24	contextual	contextual	ADJ
fcis-6912	92	25	information	information	NOUN
fcis-6912	92	26	.	.	PUNCT
fcis-6912	93	1	to	to	PART
fcis-6912	93	2	solve	solve	VERB
fcis-6912	93	3	this	this	DET
fcis-6912	93	4	problem	problem	NOUN
fcis-6912	93	5	,	,	PUNCT
fcis-6912	93	6	long	long	ADJ
fcis-6912	93	7	short	short	ADJ
fcis-6912	93	8	-	-	PUNCT
fcis-6912	93	9	term	term	NOUN
fcis-6912	93	10	memory	memory	NOUN
fcis-6912	93	11	(	(	PUNCT
fcis-6912	93	12	lstm	lstm	NOUN
fcis-6912	93	13	)	)	PUNCT
fcis-6912	93	14	uses	use	VERB
fcis-6912	93	15	gate	gate	NOUN
fcis-6912	93	16	units	unit	NOUN
fcis-6912	93	17	to	to	PART
fcis-6912	93	18	control	control	VERB
fcis-6912	93	19	the	the	DET
fcis-6912	93	20	process	process	NOUN
fcis-6912	93	21	of	of	ADP
fcis-6912	93	22	long	long	ADJ
fcis-6912	93	23	-	-	PUNCT
fcis-6912	93	24	term	term	NOUN
fcis-6912	93	25	information	information	NOUN
fcis-6912	93	26	transfer	transfer	NOUN
fcis-6912	93	27	and	and	CCONJ
fcis-6912	93	28	enhance	enhance	VERB
fcis-6912	93	29	the	the	DET
fcis-6912	93	30	correlation	correlation	NOUN
fcis-6912	93	31	of	of	ADP
fcis-6912	93	32	context	context	NOUN
fcis-6912	93	33	.	.	PUNCT
fcis-6912	94	1	lstm	lstm	NOUN
fcis-6912	94	2	has	have	VERB
fcis-6912	94	3	three	three	NUM
fcis-6912	94	4	gate	gate	NOUN
fcis-6912	94	5	structures	structure	NOUN
fcis-6912	94	6	,	,	PUNCT
fcis-6912	94	7	including	include	VERB
fcis-6912	94	8	input	input	NOUN
fcis-6912	94	9	gate	gate	NOUN
fcis-6912	94	10	(	(	PUNCT
fcis-6912	94	11	i	i	NOUN
fcis-6912	94	12	)	)	PUNCT
fcis-6912	94	13	,	,	PUNCT
fcis-6912	94	14	forget	forget	VERB
fcis-6912	94	15	gate	gate	NOUN
fcis-6912	94	16	(	(	PUNCT
fcis-6912	94	17	f	f	PROPN
fcis-6912	94	18	)	)	PUNCT
fcis-6912	94	19	,	,	PUNCT
fcis-6912	94	20	output	output	NOUN
fcis-6912	94	21	gate	gate	NOUN
fcis-6912	94	22	(	(	PUNCT
fcis-6912	94	23	o	o	NOUN
fcis-6912	94	24	)	)	PUNCT
fcis-6912	94	25	,	,	PUNCT
fcis-6912	94	26	and	and	CCONJ
fcis-6912	94	27	memory	memory	NOUN
fcis-6912	94	28	cell	cell	NOUN
fcis-6912	94	29	(	(	PUNCT
fcis-6912	94	30	c	c	NOUN
fcis-6912	94	31	)	)	PUNCT
fcis-6912	94	32	,	,	PUNCT
fcis-6912	94	33	which	which	PRON
fcis-6912	94	34	can	can	AUX
fcis-6912	94	35	effectively	effectively	ADV
fcis-6912	94	36	overcome	overcome	VERB
fcis-6912	94	37	the	the	DET
fcis-6912	94	38	problem	problem	NOUN
fcis-6912	94	39	of	of	ADP
fcis-6912	94	40	gradient	gradient	ADJ
fcis-6912	94	41	vanishing	vanishing	NOUN
fcis-6912	94	42	that	that	PRON
fcis-6912	94	43	exists	exist	VERB
fcis-6912	94	44	in	in	ADP
fcis-6912	94	45	general	general	ADJ
fcis-6912	94	46	neural	neural	ADJ
fcis-6912	94	47	networks	network	NOUN
fcis-6912	94	48	.	.	PUNCT
fcis-6912	95	1	the	the	DET
fcis-6912	95	2	structure	structure	NOUN
fcis-6912	95	3	of	of	ADP
fcis-6912	95	4	lstm	lstm	NOUN
fcis-6912	95	5	is	be	AUX
fcis-6912	95	6	shown	show	VERB
fcis-6912	95	7	in	in	ADP
fcis-6912	95	8	figure	figure	NOUN
fcis-6912	95	9	3	3	NUM
fcis-6912	95	10	.	.	PUNCT
fcis-6912	95	11	fig	fig	NOUN
fcis-6912	95	12	.	.	PUNCT
fcis-6912	96	1	3	3	NUM
fcis-6912	96	2	lstm	lstm	NOUN
fcis-6912	96	3	structure	structure	NOUN
fcis-6912	96	4	in	in	ADP
fcis-6912	96	5	figure	figure	NOUN
fcis-6912	96	6	3	3	NUM
fcis-6912	96	7	,	,	PUNCT
fcis-6912	96	8	xt	xt	X
fcis-6912	96	9	is	be	AUX
fcis-6912	96	10	the	the	DET
fcis-6912	96	11	input	input	NOUN
fcis-6912	96	12	at	at	ADP
fcis-6912	96	13	time	time	NOUN
fcis-6912	96	14	t	t	PROPN
fcis-6912	96	15	;	;	PUNCT
fcis-6912	96	16	ft	ft	X
fcis-6912	96	17	is	be	AUX
fcis-6912	96	18	the	the	DET
fcis-6912	96	19	output	output	NOUN
fcis-6912	96	20	for	for	ADP
fcis-6912	96	21	t+1	t+1	PRON
fcis-6912	96	22	at	at	ADP
fcis-6912	96	23	time	time	NOUN
fcis-6912	96	24	t	t	PROPN
fcis-6912	96	25	;	;	PUNCT
fcis-6912	96	26	ot	ot	X
fcis-6912	96	27	is	be	AUX
fcis-6912	96	28	the	the	DET
fcis-6912	96	29	output	output	NOUN
fcis-6912	96	30	at	at	ADP
fcis-6912	96	31	time	time	NOUN
fcis-6912	96	32	t	t	PROPN
fcis-6912	96	33	;	;	PUNCT
fcis-6912	96	34	ht	ht	PROPN
fcis-6912	96	35	is	be	AUX
fcis-6912	96	36	the	the	DET
fcis-6912	96	37	hidden	hidden	ADJ
fcis-6912	96	38	layer	layer	NOUN
fcis-6912	96	39	representing	represent	VERB
fcis-6912	96	40	the	the	DET
fcis-6912	96	41	output	output	NOUN
fcis-6912	96	42	at	at	ADP
fcis-6912	96	43	time	time	NOUN
fcis-6912	96	44	t	t	PROPN
fcis-6912	96	45	;	;	PUNCT
fcis-6912	96	46	σ	σ	X
fcis-6912	96	47	is	be	AUX
fcis-6912	96	48	the	the	DET
fcis-6912	96	49	sigmoid	sigmoid	NOUN
fcis-6912	96	50	function	function	NOUN
fcis-6912	96	51	;	;	PUNCT
fcis-6912	96	52	ct	ct	PROPN
fcis-6912	96	53	is	be	AUX
fcis-6912	96	54	the	the	DET
fcis-6912	96	55	cell	cell	NOUN
fcis-6912	96	56	state	state	NOUN
fcis-6912	96	57	at	at	ADP
fcis-6912	96	58	time	time	NOUN
fcis-6912	96	59	t.	t.	PROPN
fcis-6912	96	60	the	the	DET
fcis-6912	96	61	gate	gate	PROPN
fcis-6912	96	62	unit	unit	NOUN
fcis-6912	96	63	calculation	calculation	NOUN
fcis-6912	96	64	formulas	formula	NOUN
fcis-6912	96	65	in	in	ADP
fcis-6912	96	66	lstm	lstm	NOUN
fcis-6912	96	67	are	be	AUX
fcis-6912	96	68	shown	show	VERB
fcis-6912	96	69	below	below	ADP
fcis-6912	96	70	:	:	PUNCT
fcis-6912	96	71	𝑖𝑡	𝑖𝑡	PROPN
fcis-6912	96	72	=	=	PUNCT
fcis-6912	96	73	𝜎(𝑤𝑖	𝜎(𝑤𝑖	NUM
fcis-6912	96	74	⋅	⋅	PROPN
fcis-6912	97	1	[	[	X
fcis-6912	97	2	ℎ𝑡	ℎ𝑡	NOUN
fcis-6912	97	3	,	,	PUNCT
fcis-6912	97	4	𝑥𝑡	𝑥𝑡	ADP
fcis-6912	97	5	]	]	X
fcis-6912	97	6	+	+	CCONJ
fcis-6912	97	7	𝑏𝑖	𝑏𝑖	X
fcis-6912	97	8	)	)	PUNCT
fcis-6912	97	9	(	(	PUNCT
fcis-6912	97	10	1	1	X
fcis-6912	97	11	)	)	PUNCT
fcis-6912	97	12	𝑓𝑡	𝑓𝑡	NOUN
fcis-6912	97	13	=	=	PUNCT
fcis-6912	97	14	𝜎(𝑤𝑓	𝜎(𝑤𝑓	NUM
fcis-6912	97	15	⋅	⋅	X
fcis-6912	97	16	[	[	X
fcis-6912	97	17	ℎ𝑡	ℎ𝑡	X
fcis-6912	97	18	,	,	PUNCT
fcis-6912	97	19	𝑥𝑡	𝑥𝑡	PROPN
fcis-6912	97	20	]	]	PUNCT
fcis-6912	97	21	+	+	CCONJ
fcis-6912	97	22	𝑏𝑓	𝑏𝑓	PROPN
fcis-6912	97	23	)	)	PUNCT
fcis-6912	97	24	(	(	PUNCT
fcis-6912	97	25	2	2	X
fcis-6912	97	26	)	)	PUNCT
fcis-6912	97	27	𝑜𝑡	𝑜𝑡	NOUN
fcis-6912	97	28	=	=	SYM
fcis-6912	97	29	𝜎(𝑤𝑜	𝜎(𝑤𝑜	NUM
fcis-6912	97	30	⋅	⋅	PROPN
fcis-6912	97	31	[	[	X
fcis-6912	97	32	ℎ𝑡	ℎ𝑡	NOUN
fcis-6912	97	33	,	,	PUNCT
fcis-6912	97	34	𝑥𝑡	𝑥𝑡	ADP
fcis-6912	97	35	]	]	X
fcis-6912	97	36	+	+	CCONJ
fcis-6912	97	37	𝑏𝑜	𝑏𝑜	X
fcis-6912	97	38	)	)	PUNCT
fcis-6912	97	39	(	(	PUNCT
fcis-6912	97	40	3	3	X
fcis-6912	97	41	)	)	PUNCT
fcis-6912	97	42	ℎ𝑡	ℎ𝑡	NOUN
fcis-6912	97	43	=	=	PUNCT
fcis-6912	98	1	𝑂𝑡	𝑂𝑡	PROPN
fcis-6912	98	2	tanh	tanh	PROPN
fcis-6912	98	3	𝐶𝑡	𝐶𝑡	PROPN
fcis-6912	98	4	(	(	PUNCT
fcis-6912	98	5	4	4	NUM
fcis-6912	98	6	)	)	PUNCT
fcis-6912	98	7	𝑐𝑡	𝑐𝑡	NOUN
fcis-6912	98	8	=	=	PUNCT
fcis-6912	98	9	𝑓𝑡𝐶𝑡−1	𝑓𝑡𝐶𝑡−1	VERB
fcis-6912	99	1	+	+	CCONJ
fcis-6912	100	1	𝑖𝑡tanh(𝑤𝑐	𝑖𝑡tanh(𝑤𝑐	ADJ
fcis-6912	100	2	⋅	⋅	PROPN
fcis-6912	100	3	[	[	X
fcis-6912	100	4	ht−1	ht−1	PROPN
fcis-6912	100	5	,	,	PUNCT
fcis-6912	100	6	xt	xt	ADP
fcis-6912	100	7	]	]	X
fcis-6912	100	8	+	+	CCONJ
fcis-6912	100	9	bc	bc	PROPN
fcis-6912	100	10	)	)	PUNCT
fcis-6912	100	11	(	(	PUNCT
fcis-6912	100	12	5	5	NUM
fcis-6912	100	13	)	)	PUNCT
fcis-6912	100	14	in	in	ADP
fcis-6912	100	15	the	the	DET
fcis-6912	100	16	formula	formula	NOUN
fcis-6912	100	17	,	,	PUNCT
fcis-6912	100	18	wi	wi	PROPN
fcis-6912	100	19	,	,	PUNCT
fcis-6912	100	20	wf	wf	PROPN
fcis-6912	100	21	,	,	PUNCT
fcis-6912	100	22	and	and	CCONJ
fcis-6912	100	23	wo	will	AUX
fcis-6912	100	24	are	be	AUX
fcis-6912	100	25	weight	weight	NOUN
fcis-6912	100	26	matrices	matrix	NOUN
fcis-6912	100	27	that	that	PRON
fcis-6912	100	28	connect	connect	VERB
fcis-6912	100	29	to	to	ADP
fcis-6912	100	30	the	the	DET
fcis-6912	100	31	gate	gate	NOUN
fcis-6912	100	32	unit	unit	NOUN
fcis-6912	100	33	;	;	PUNCT
fcis-6912	100	34	bi	bi	NOUN
fcis-6912	100	35	,	,	PUNCT
fcis-6912	100	36	bf	bf	NOUN
fcis-6912	100	37	,	,	PUNCT
fcis-6912	100	38	and	and	CCONJ
fcis-6912	100	39	bo	bo	PROPN
fcis-6912	100	40	are	be	AUX
fcis-6912	100	41	the	the	DET
fcis-6912	100	42	bias	bias	NOUN
fcis-6912	100	43	values	value	NOUN
fcis-6912	100	44	.	.	PUNCT
fcis-6912	101	1	4	4	X
fcis-6912	101	2	.	.	X
fcis-6912	101	3	experiments	experiment	NOUN
fcis-6912	101	4	and	and	CCONJ
fcis-6912	101	5	results	result	VERB
fcis-6912	101	6	4.1	4.1	NUM
fcis-6912	101	7	.	.	PUNCT
fcis-6912	101	8	experimental	experimental	ADJ
fcis-6912	101	9	environment	environment	NOUN
fcis-6912	101	10	and	and	CCONJ
fcis-6912	101	11	experimental	experimental	ADJ
fcis-6912	101	12	data	datum	NOUN
fcis-6912	101	13	set	set	VERB
fcis-6912	101	14	the	the	DET
fcis-6912	101	15	specific	specific	ADJ
fcis-6912	101	16	experimental	experimental	ADJ
fcis-6912	101	17	environment	environment	NOUN
fcis-6912	101	18	is	be	AUX
fcis-6912	101	19	as	as	SCONJ
fcis-6912	101	20	follows	follow	VERB
fcis-6912	101	21	:	:	PUNCT
fcis-6912	101	22	python	python	PROPN
fcis-6912	101	23	was	be	AUX
fcis-6912	101	24	chosen	choose	VERB
fcis-6912	101	25	as	as	ADP
fcis-6912	101	26	the	the	DET
fcis-6912	101	27	development	development	NOUN
fcis-6912	101	28	language	language	NOUN
fcis-6912	101	29	,	,	PUNCT
fcis-6912	101	30	the	the	DET
fcis-6912	101	31	cpu	cpu	NOUN
fcis-6912	101	32	model	model	NOUN
fcis-6912	101	33	is	be	AUX
fcis-6912	101	34	amd	amd	ADJ
fcis-6912	101	35	ryzen	ryzen	ADJ
fcis-6912	101	36	7	7	NUM
fcis-6912	101	37	4800h	4800h	NUM
fcis-6912	101	38	,	,	PUNCT
fcis-6912	101	39	and	and	CCONJ
fcis-6912	101	40	visual	visual	ADJ
fcis-6912	101	41	studio	studio	NOUN
fcis-6912	101	42	code	code	NOUN
fcis-6912	101	43	was	be	AUX
fcis-6912	101	44	chosen	choose	VERB
fcis-6912	101	45	as	as	ADP
fcis-6912	101	46	the	the	DET
fcis-6912	101	47	development	development	NOUN
fcis-6912	101	48	tool	tool	NOUN
fcis-6912	101	49	.	.	PUNCT
fcis-6912	102	1	the	the	DET
fcis-6912	102	2	experiment	experiment	NOUN
fcis-6912	102	3	used	use	VERB
fcis-6912	102	4	the	the	DET
fcis-6912	102	5	people	people	NOUN
fcis-6912	102	6	's	's	PART
fcis-6912	102	7	daily	daily	ADJ
fcis-6912	102	8	corpus	corpus	NOUN
fcis-6912	102	9	from	from	ADP
fcis-6912	102	10	january	january	PROPN
fcis-6912	102	11	1998	1998	NUM
fcis-6912	102	12	as	as	ADP
fcis-6912	102	13	the	the	DET
fcis-6912	102	14	dataset	dataset	NOUN
fcis-6912	102	15	.	.	PUNCT
fcis-6912	103	1	this	this	DET
fcis-6912	103	2	corpus	corpus	NOUN
fcis-6912	103	3	is	be	AUX
fcis-6912	103	4	a	a	DET
fcis-6912	103	5	tagged	tag	VERB
fcis-6912	103	6	corpus	corpus	NOUN
fcis-6912	103	7	produced	produce	VERB
fcis-6912	103	8	jointly	jointly	ADV
fcis-6912	103	9	by	by	ADP
fcis-6912	103	10	the	the	DET
fcis-6912	103	11	peking	peking	PROPN
fcis-6912	103	12	university	university	PROPN
fcis-6912	103	13	computational	computational	ADJ
fcis-6912	103	14	linguistics	linguistic	NOUN
fcis-6912	103	15	laboratory	laboratory	NOUN
fcis-6912	103	16	and	and	CCONJ
fcis-6912	103	17	the	the	DET
fcis-6912	103	18	fujitsu	fujitsu	PROPN
fcis-6912	103	19	research	research	NOUN
fcis-6912	103	20	center	center	NOUN
fcis-6912	103	21	,	,	PUNCT
fcis-6912	103	22	and	and	CCONJ
fcis-6912	103	23	has	have	AUX
fcis-6912	103	24	been	be	AUX
fcis-6912	103	25	used	use	VERB
fcis-6912	103	26	as	as	ADP
fcis-6912	103	27	raw	raw	ADJ
fcis-6912	103	28	data	datum	NOUN
fcis-6912	103	29	in	in	ADP
fcis-6912	103	30	a	a	DET
fcis-6912	103	31	large	large	ADJ
fcis-6912	103	32	number	number	NOUN
fcis-6912	103	33	of	of	ADP
fcis-6912	103	34	research	research	NOUN
fcis-6912	103	35	studies	study	NOUN
fcis-6912	103	36	and	and	CCONJ
fcis-6912	103	37	papers	paper	NOUN
fcis-6912	103	38	.	.	PUNCT
fcis-6912	104	1	the	the	DET
fcis-6912	104	2	experimental	experimental	ADJ
fcis-6912	104	3	process	process	NOUN
fcis-6912	104	4	involved	involve	VERB
fcis-6912	104	5	randomly	randomly	ADV
fcis-6912	104	6	dividing	divide	VERB
fcis-6912	104	7	the	the	DET
fcis-6912	104	8	dataset	dataset	NOUN
fcis-6912	104	9	into	into	ADP
fcis-6912	104	10	training	training	NOUN
fcis-6912	104	11	set	set	NOUN
fcis-6912	104	12	,	,	PUNCT
fcis-6912	104	13	validation	validation	NOUN
fcis-6912	104	14	set	set	NOUN
fcis-6912	104	15	,	,	PUNCT
fcis-6912	104	16	and	and	CCONJ
fcis-6912	104	17	test	test	NOUN
fcis-6912	104	18	set	set	VERB
fcis-6912	104	19	,	,	PUNCT
fcis-6912	104	20	with	with	ADP
fcis-6912	104	21	80	80	NUM
fcis-6912	104	22	%	%	NOUN
fcis-6912	104	23	,	,	PUNCT
fcis-6912	104	24	10	10	NUM
fcis-6912	104	25	%	%	NOUN
fcis-6912	104	26	,	,	PUNCT
fcis-6912	104	27	and	and	CCONJ
fcis-6912	104	28	10	10	NUM
fcis-6912	104	29	%	%	NOUN
fcis-6912	104	30	respectively	respectively	ADV
fcis-6912	104	31	.	.	PUNCT
fcis-6912	105	1	the	the	DET
fcis-6912	105	2	dataset	dataset	NOUN
fcis-6912	105	3	was	be	AUX
fcis-6912	105	4	annotated	annotate	VERB
fcis-6912	105	5	using	use	VERB
fcis-6912	105	6	the	the	DET
fcis-6912	105	7	bio	bio	NOUN
fcis-6912	105	8	tagging	tagging	NOUN
fcis-6912	105	9	scheme	scheme	NOUN
fcis-6912	105	10	(	(	PUNCT
fcis-6912	105	11	beginning	beginning	NOUN
fcis-6912	105	12	,	,	PUNCT
fcis-6912	105	13	inside	inside	ADV
fcis-6912	105	14	,	,	PUNCT
fcis-6912	105	15	outside	outside	ADV
fcis-6912	105	16	)	)	PUNCT
fcis-6912	105	17	,	,	PUNCT
fcis-6912	105	18	which	which	PRON
fcis-6912	105	19	includes	include	VERB
fcis-6912	105	20	three	three	NUM
fcis-6912	105	21	entity	entity	NOUN
fcis-6912	105	22	types	type	NOUN
fcis-6912	105	23	:	:	PUNCT
fcis-6912	105	24	person	person	NOUN
fcis-6912	105	25	names	name	NOUN
fcis-6912	105	26	(	(	PUNCT
fcis-6912	105	27	per	per	ADP
fcis-6912	105	28	)	)	PUNCT
fcis-6912	105	29	,	,	PUNCT
fcis-6912	105	30	location	location	NOUN
fcis-6912	105	31	names	name	NOUN
fcis-6912	105	32	(	(	PUNCT
fcis-6912	105	33	loc	loc	NOUN
fcis-6912	105	34	)	)	PUNCT
fcis-6912	105	35	,	,	PUNCT
fcis-6912	105	36	and	and	CCONJ
fcis-6912	105	37	organization	organization	NOUN
fcis-6912	105	38	names	name	NOUN
fcis-6912	105	39	(	(	PUNCT
fcis-6912	105	40	org	org	NOUN
fcis-6912	105	41	)	)	PUNCT
fcis-6912	105	42	,	,	PUNCT
fcis-6912	105	43	with	with	ADP
fcis-6912	105	44	a	a	DET
fcis-6912	105	45	total	total	NOUN
fcis-6912	105	46	of	of	ADP
fcis-6912	105	47	seven	seven	NUM
fcis-6912	105	48	tags	tag	NOUN
fcis-6912	105	49	(	(	PUNCT
fcis-6912	105	50	'	'	PUNCT
fcis-6912	105	51	bper	bper	NOUN
fcis-6912	105	52	'	'	PUNCT
fcis-6912	105	53	,	,	PUNCT
fcis-6912	105	54	'	'	PUNCT
fcis-6912	105	55	i	i	PRON
fcis-6912	105	56	-	-	PUNCT
fcis-6912	105	57	per	per	ADP
fcis-6912	105	58	'	'	PUNCT
fcis-6912	105	59	,	,	PUNCT
fcis-6912	105	60	'	'	PUNCT
fcis-6912	105	61	b	b	X
fcis-6912	105	62	-	-	PUNCT
fcis-6912	105	63	loc	loc	NOUN
fcis-6912	105	64	'	'	PUNCT
fcis-6912	105	65	,	,	PUNCT
fcis-6912	105	66	'	'	PUNCT
fcis-6912	105	67	i	i	PROPN
fcis-6912	105	68	-	-	PUNCT
fcis-6912	105	69	loc	loc	NOUN
fcis-6912	105	70	'	'	PUNCT
fcis-6912	105	71	,	,	PUNCT
fcis-6912	105	72	'	'	PUNCT
fcis-6912	105	73	b	b	X
fcis-6912	105	74	-	-	PUNCT
fcis-6912	105	75	org	org	NOUN
fcis-6912	105	76	'	'	PUNCT
fcis-6912	105	77	,	,	PUNCT
fcis-6912	105	78	'	'	PUNCT
fcis-6912	105	79	i	i	NOUN
fcis-6912	105	80	-	-	PUNCT
fcis-6912	105	81	org	org	NOUN
fcis-6912	105	82	'	'	PUNCT
fcis-6912	105	83	,	,	PUNCT
fcis-6912	105	84	'	'	PUNCT
fcis-6912	105	85	o	o	NOUN
fcis-6912	105	86	'	'	PUNCT
fcis-6912	105	87	)	)	PUNCT
fcis-6912	105	88	.	.	PUNCT
fcis-6912	106	1	precision	precision	NOUN
fcis-6912	106	2	(	(	PUNCT
fcis-6912	106	3	p	p	NOUN
fcis-6912	106	4	)	)	PUNCT
fcis-6912	106	5	,	,	PUNCT
fcis-6912	106	6	recall	recall	INTJ
fcis-6912	106	7	(	(	PUNCT
fcis-6912	106	8	r	r	NOUN
fcis-6912	106	9	)	)	PUNCT
fcis-6912	106	10	,	,	PUNCT
fcis-6912	106	11	and	and	CCONJ
fcis-6912	106	12	f1	f1	NOUN
fcis-6912	106	13	-	-	PUNCT
fcis-6912	106	14	score	score	NOUN
fcis-6912	106	15	were	be	AUX
fcis-6912	106	16	used	use	VERB
fcis-6912	106	17	as	as	ADP
fcis-6912	106	18	evaluation	evaluation	NOUN
fcis-6912	106	19	metrics	metric	NOUN
fcis-6912	106	20	for	for	ADP
fcis-6912	106	21	the	the	DET
fcis-6912	106	22	model	model	NOUN
fcis-6912	106	23	in	in	ADP
fcis-6912	106	24	the	the	DET
fcis-6912	106	25	experiment	experiment	NOUN
fcis-6912	106	26	,	,	PUNCT
fcis-6912	106	27	and	and	CCONJ
fcis-6912	106	28	were	be	AUX
fcis-6912	106	29	calculated	calculate	VERB
fcis-6912	106	30	using	use	VERB
fcis-6912	106	31	the	the	DET
fcis-6912	106	32	following	follow	VERB
fcis-6912	106	33	formulas	formula	NOUN
fcis-6912	106	34	:	:	PUNCT
fcis-6912	106	35	𝑃	𝑃	NOUN
fcis-6912	106	36	=	=	PUNCT
fcis-6912	106	37	模型正确标注的实体个数	模型正确标注的实体个数	NOUN
fcis-6912	106	38	模型标注的所有实体个数	模型标注的所有实体个数	ADJ
fcis-6912	106	39	×	×	NOUN
fcis-6912	106	40	100	100	NUM
fcis-6912	106	41	%	%	NOUN
fcis-6912	106	42	(	(	PUNCT
fcis-6912	106	43	6	6	NUM
fcis-6912	106	44	)	)	PUNCT
fcis-6912	106	45	𝑅	𝑅	NOUN
fcis-6912	106	46	=	=	PUNCT
fcis-6912	106	47	模型正确标注的实体个数	模型正确标注的实体个数	NOUN
fcis-6912	107	1	样本中所有实体个数	样本中所有实体个数	ADJ
fcis-6912	107	2	×	×	PROPN
fcis-6912	107	3	100	100	NUM
fcis-6912	107	4	%	%	NOUN
fcis-6912	107	5	(	(	PUNCT
fcis-6912	107	6	7	7	NUM
fcis-6912	107	7	)	)	PUNCT
fcis-6912	107	8	𝐹	𝐹	PROPN
fcis-6912	107	9	=	=	SYM
fcis-6912	107	10	2×𝑃×𝑅	2×𝑃×𝑅	NOUN
fcis-6912	107	11	𝑃+𝑅	𝑃+𝑅	NOUN
fcis-6912	107	12	×	×	NOUN
fcis-6912	107	13	100	100	NUM
fcis-6912	107	14	%	%	NOUN
fcis-6912	107	15	(	(	PUNCT
fcis-6912	107	16	8)	8)	NUM
fcis-6912	107	17	4.2	4.2	NUM
fcis-6912	107	18	.	.	PUNCT
fcis-6912	108	1	experimental	experimental	ADJ
fcis-6912	108	2	result	result	NOUN
fcis-6912	108	3	the	the	DET
fcis-6912	108	4	experiment	experiment	NOUN
fcis-6912	108	5	compares	compare	VERB
fcis-6912	108	6	the	the	DET
fcis-6912	108	7	model	model	NOUN
fcis-6912	108	8	in	in	ADP
fcis-6912	108	9	the	the	DET
fcis-6912	108	10	article	article	NOUN
fcis-6912	108	11	with	with	ADP
fcis-6912	108	12	other	other	ADJ
fcis-6912	108	13	models	model	NOUN
fcis-6912	108	14	to	to	PART
fcis-6912	108	15	validate	validate	VERB
fcis-6912	108	16	its	its	PRON
fcis-6912	108	17	effectiveness	effectiveness	NOUN
fcis-6912	108	18	.	.	PUNCT
fcis-6912	109	1	several	several	ADJ
fcis-6912	109	2	models	model	NOUN
fcis-6912	109	3	with	with	ADP
fcis-6912	109	4	good	good	ADJ
fcis-6912	109	5	performance	performance	NOUN
fcis-6912	109	6	and	and	CCONJ
fcis-6912	109	7	mainstream	mainstream	NOUN
fcis-6912	109	8	popularity	popularity	NOUN
fcis-6912	109	9	were	be	AUX
fcis-6912	109	10	selected	select	VERB
fcis-6912	109	11	for	for	ADP
fcis-6912	109	12	comparison	comparison	NOUN
fcis-6912	109	13	with	with	ADP
fcis-6912	109	14	the	the	DET
fcis-6912	109	15	model	model	NOUN
fcis-6912	109	16	in	in	ADP
fcis-6912	109	17	the	the	DET
fcis-6912	109	18	article	article	NOUN
fcis-6912	109	19	,	,	PUNCT
fcis-6912	109	20	including	include	VERB
fcis-6912	109	21	crf	crf	PROPN
fcis-6912	109	22	,	,	PUNCT
fcis-6912	109	23	lstm	lstm	PROPN
fcis-6912	109	24	-	-	PUNCT
fcis-6912	109	25	crf	crf	NOUN
fcis-6912	109	26	,	,	PUNCT
fcis-6912	109	27	and	and	CCONJ
fcis-6912	109	28	gru	gru	PROPN
fcis-6912	109	29	-	-	PROPN
fcis-6912	109	30	crf	crf	NOUN
fcis-6912	109	31	,	,	PUNCT
fcis-6912	109	32	and	and	CCONJ
fcis-6912	109	33	the	the	DET
fcis-6912	109	34	results	result	NOUN
fcis-6912	109	35	are	be	AUX
fcis-6912	109	36	shown	show	VERB
fcis-6912	109	37	in	in	ADP
fcis-6912	109	38	table	table	NOUN
fcis-6912	109	39	1	1	NUM
fcis-6912	109	40	.	.	PUNCT
fcis-6912	110	1	the	the	DET
fcis-6912	110	2	experimental	experimental	ADJ
fcis-6912	110	3	results	result	NOUN
fcis-6912	110	4	show	show	VERB
fcis-6912	110	5	that	that	SCONJ
fcis-6912	110	6	the	the	DET
fcis-6912	110	7	model	model	NOUN
fcis-6912	110	8	in	in	ADP
fcis-6912	110	9	the	the	DET
fcis-6912	110	10	article	article	NOUN
fcis-6912	110	11	has	have	AUX
fcis-6912	110	12	improved	improve	VERB
fcis-6912	110	13	in	in	ADP
fcis-6912	110	14	all	all	DET
fcis-6912	110	15	metrics	metric	NOUN
fcis-6912	110	16	compared	compare	VERB
fcis-6912	110	17	to	to	ADP
fcis-6912	110	18	other	other	ADJ
fcis-6912	110	19	models	model	NOUN
fcis-6912	110	20	.	.	PUNCT
fcis-6912	111	1	by	by	ADP
fcis-6912	111	2	comparing	compare	VERB
fcis-6912	111	3	the	the	DET
fcis-6912	111	4	experimental	experimental	ADJ
fcis-6912	111	5	results	result	NOUN
fcis-6912	111	6	on	on	ADP
fcis-6912	111	7	the	the	DET
fcis-6912	111	8	same	same	ADJ
fcis-6912	111	9	dataset	dataset	NOUN
fcis-6912	111	10	,	,	PUNCT
fcis-6912	111	11	it	it	PRON
fcis-6912	111	12	is	be	AUX
fcis-6912	111	13	demonstrated	demonstrate	VERB
fcis-6912	111	14	that	that	SCONJ
fcis-6912	111	15	applying	apply	VERB
fcis-6912	111	16	the	the	DET
fcis-6912	111	17	model	model	NOUN
fcis-6912	111	18	in	in	ADP
fcis-6912	111	19	the	the	DET
fcis-6912	111	20	article	article	NOUN
fcis-6912	111	21	can	can	AUX
fcis-6912	111	22	improve	improve	VERB
fcis-6912	111	23	the	the	DET
fcis-6912	111	24	recognition	recognition	NOUN
fcis-6912	111	25	performance	performance	NOUN
fcis-6912	111	26	in	in	ADP
fcis-6912	111	27	entity	entity	NOUN
fcis-6912	111	28	recognition	recognition	NOUN
fcis-6912	111	29	tasks	task	NOUN
fcis-6912	111	30	.	.	PUNCT
fcis-6912	112	1	table.1	table.1	VERB
fcis-6912	112	2	comparison	comparison	NOUN
fcis-6912	112	3	of	of	ADP
fcis-6912	112	4	experimental	experimental	ADJ
fcis-6912	112	5	results	result	NOUN
fcis-6912	112	6	of	of	ADP
fcis-6912	112	7	different	different	ADJ
fcis-6912	112	8	models	model	NOUN
fcis-6912	112	9	%	%	INTJ
fcis-6912	113	1	5	5	NUM
fcis-6912	113	2	.	.	PUNCT
fcis-6912	113	3	conclusion	conclusion	NOUN
fcis-6912	113	4	to	to	PART
fcis-6912	113	5	address	address	VERB
fcis-6912	113	6	the	the	DET
fcis-6912	113	7	problem	problem	NOUN
fcis-6912	113	8	that	that	SCONJ
fcis-6912	113	9	traditional	traditional	ADJ
fcis-6912	113	10	word	word	NOUN
fcis-6912	113	11	vectors	vector	NOUN
fcis-6912	113	12	can	can	AUX
fcis-6912	113	13	not	not	PART
fcis-6912	113	14	represent	represent	VERB
fcis-6912	113	15	the	the	DET
fcis-6912	113	16	polysemy	polysemy	NOUN
fcis-6912	113	17	of	of	ADP
fcis-6912	113	18	characters	character	NOUN
fcis-6912	113	19	and	and	CCONJ
fcis-6912	113	20	models	model	NOUN
fcis-6912	113	21	have	have	VERB
fcis-6912	113	22	difficulty	difficulty	NOUN
fcis-6912	113	23	in	in	ADP
fcis-6912	113	24	obtaining	obtain	VERB
fcis-6912	113	25	complete	complete	ADJ
fcis-6912	113	26	semantic	semantic	ADJ
fcis-6912	113	27	representations	representation	NOUN
fcis-6912	113	28	,	,	PUNCT
fcis-6912	113	29	this	this	DET
fcis-6912	113	30	paper	paper	NOUN
fcis-6912	113	31	proposes	propose	VERB
fcis-6912	113	32	the	the	DET
fcis-6912	113	33	ernie	ernie	NOUN
fcis-6912	113	34	-	-	PUNCT
fcis-6912	113	35	bigru	bigru	PROPN
fcis-6912	113	36	-	-	PUNCT
fcis-6912	113	37	crf	crf	NOUN
fcis-6912	113	38	model	model	NOUN
fcis-6912	113	39	.	.	PUNCT
fcis-6912	114	1	the	the	DET
fcis-6912	114	2	model	model	NOUN
fcis-6912	114	3	uses	use	VERB
fcis-6912	114	4	multi	multi	ADJ
fcis-6912	114	5	-	-	ADJ
fcis-6912	114	6	layer	layer	ADJ
fcis-6912	114	7	bidirectional	bidirectional	ADJ
fcis-6912	114	8	transformers	transformer	NOUN
fcis-6912	114	9	as	as	SCONJ
fcis-6912	114	10	encoders	encoder	NOUN
fcis-6912	114	11	to	to	PART
fcis-6912	114	12	extract	extract	VERB
fcis-6912	114	13	features	feature	NOUN
fcis-6912	114	14	,	,	PUNCT
fcis-6912	114	15	and	and	CCONJ
fcis-6912	114	16	adopts	adopt	VERB
fcis-6912	114	17	three	three	NUM
fcis-6912	114	18	levels	level	NOUN
fcis-6912	114	19	of	of	ADP
fcis-6912	114	20	masking	masking	NOUN
fcis-6912	114	21	strategies	strategy	NOUN
fcis-6912	114	22	including	include	VERB
fcis-6912	114	23	character	character	NOUN
fcis-6912	114	24	masking	masking	NOUN
fcis-6912	114	25	,	,	PUNCT
fcis-6912	114	26	phrase	phrase	NOUN
fcis-6912	114	27	masking	masking	NOUN
fcis-6912	114	28	,	,	PUNCT
fcis-6912	114	29	and	and	CCONJ
fcis-6912	114	30	entity	entity	NOUN
fcis-6912	114	31	masking	mask	VERB
fcis-6912	114	32	to	to	PART
fcis-6912	114	33	dynamically	dynamically	ADV
fcis-6912	114	34	generate	generate	VERB
fcis-6912	114	35	context	context	NOUN
fcis-6912	114	36	semantic	semantic	ADJ
fcis-6912	114	37	representations	representation	NOUN
fcis-6912	114	38	of	of	ADP
fcis-6912	114	39	characters	character	NOUN
fcis-6912	114	40	.	.	PUNCT
fcis-6912	115	1	compared	compare	VERB
fcis-6912	115	2	to	to	ADP
fcis-6912	115	3	traditional	traditional	ADJ
fcis-6912	115	4	word	word	NOUN
fcis-6912	115	5	vectors	vector	NOUN
fcis-6912	115	6	,	,	PUNCT
fcis-6912	115	7	ernie	ernie	NOUN
fcis-6912	115	8	-	-	PUNCT
fcis-6912	115	9	bigru	bigru	PROPN
fcis-6912	115	10	-	-	PUNCT
fcis-6912	115	11	crf	crf	NOUN
fcis-6912	115	12	can	can	AUX
fcis-6912	115	13	enhance	enhance	VERB
fcis-6912	115	14	the	the	DET
fcis-6912	115	15	model	model	NOUN
fcis-6912	115	16	's	's	PART
fcis-6912	115	17	semantic	semantic	ADJ
fcis-6912	115	18	representation	representation	NOUN
fcis-6912	115	19	ability	ability	NOUN
fcis-6912	115	20	and	and	CCONJ
fcis-6912	115	21	improve	improve	VERB
fcis-6912	115	22	named	name	VERB
fcis-6912	115	23	entity	entity	NOUN
fcis-6912	115	24	recognition	recognition	NOUN
fcis-6912	115	25	performance	performance	NOUN
fcis-6912	115	26	.	.	PUNCT
fcis-6912	116	1	however	however	ADV
fcis-6912	116	2	,	,	PUNCT
fcis-6912	116	3	in	in	ADP
fcis-6912	116	4	specific	specific	ADJ
fcis-6912	116	5	domains	domain	NOUN
fcis-6912	116	6	that	that	PRON
fcis-6912	116	7	lack	lack	VERB
fcis-6912	116	8	large	large	ADJ
fcis-6912	116	9	-	-	PUNCT
fcis-6912	116	10	scale	scale	NOUN
fcis-6912	116	11	annotated	annotate	VERB
fcis-6912	116	12	data	datum	NOUN
fcis-6912	116	13	,	,	PUNCT
fcis-6912	116	14	the	the	DET
fcis-6912	116	15	model	model	NOUN
fcis-6912	116	16	may	may	AUX
fcis-6912	116	17	extract	extract	VERB
fcis-6912	116	18	incorrectly	incorrectly	ADV
fcis-6912	116	19	due	due	ADJ
fcis-6912	116	20	to	to	ADP
fcis-6912	116	21	insufficient	insufficient	ADJ
fcis-6912	116	22	context	context	NOUN
fcis-6912	116	23	information	information	NOUN
fcis-6912	116	24	and	and	CCONJ
fcis-6912	116	25	the	the	DET
fcis-6912	116	26	presence	presence	NOUN
fcis-6912	116	27	of	of	ADP
fcis-6912	116	28	abbreviations	abbreviation	NOUN
fcis-6912	116	29	and	and	CCONJ
fcis-6912	116	30	ambiguous	ambiguous	ADJ
fcis-6912	116	31	entities	entity	NOUN
fcis-6912	116	32	.	.	PUNCT
fcis-6912	117	1	therefore	therefore	ADV
fcis-6912	117	2	,	,	PUNCT
fcis-6912	117	3	the	the	DET
fcis-6912	117	4	next	next	ADJ
fcis-6912	117	5	research	research	NOUN
fcis-6912	117	6	direction	direction	NOUN
fcis-6912	117	7	can	can	AUX
fcis-6912	117	8	consider	consider	VERB
fcis-6912	117	9	combining	combine	VERB
fcis-6912	117	10	deep	deep	ADJ
fcis-6912	117	11	24	24	NUM
fcis-6912	117	12	learning	learn	VERB
fcis-6912	117	13	with	with	ADP
fcis-6912	117	14	transfer	transfer	NOUN
fcis-6912	117	15	learning	learning	NOUN
fcis-6912	117	16	methods	method	NOUN
fcis-6912	117	17	to	to	PART
fcis-6912	117	18	address	address	VERB
fcis-6912	117	19	these	these	DET
fcis-6912	117	20	issues	issue	NOUN
fcis-6912	117	21	.	.	PUNCT
fcis-6912	118	1	references	reference	NOUN
fcis-6912	118	2	[	[	X
fcis-6912	118	3	1	1	NUM
fcis-6912	118	4	]	]	PUNCT
fcis-6912	118	5	marrero	marrero	PROPN
fcis-6912	118	6	m	m	PROPN
fcis-6912	118	7	,	,	PUNCT
fcis-6912	118	8	urbano	urbano	PROPN
fcis-6912	118	9	j	j	PROPN
fcis-6912	118	10	,	,	PUNCT
fcis-6912	118	11	s	s	PART
fcis-6912	118	12	á	á	PROPN
fcis-6912	118	13	nchez	nchez	PROPN
fcis-6912	118	14	-	-	PUNCT
fcis-6912	118	15	uadrado	uadrado	PROPN
fcis-6912	118	16	s	s	PROPN
fcis-6912	118	17	,	,	PUNCT
fcis-6912	118	18	et	et	PROPN
fcis-6912	118	19	al	al	PROPN
fcis-6912	118	20	.	.	PROPN
fcis-6912	118	21	named	name	VERB
fcis-6912	118	22	entity	entity	NOUN
fcis-6912	118	23	recognition	recognition	NOUN
fcis-6912	118	24	:	:	PUNCT
fcis-6912	118	25	fallacies	fallacy	NOUN
fcis-6912	118	26	,	,	PUNCT
fcis-6912	118	27	challenges	challenge	NOUN
fcis-6912	118	28	,	,	PUNCT
fcis-6912	118	29	and	and	CCONJ
fcis-6912	118	30	opportunities	opportunity	NOUN
fcis-6912	119	1	[	[	X
fcis-6912	119	2	j	j	X
fcis-6912	119	3	]	]	X
fcis-6912	119	4	.	.	PUNCT
fcis-6912	120	1	computer	computer	NOUN
fcis-6912	120	2	standards	standard	NOUN
fcis-6912	120	3	and	and	CCONJ
fcis-6912	120	4	interfaces	interface	NOUN
fcis-6912	120	5	,	,	PUNCT
fcis-6912	120	6	2013	2013	NUM
fcis-6912	120	7	,	,	PUNCT
fcis-6912	120	8	35	35	NUM
fcis-6912	120	9	(	(	PUNCT
fcis-6912	120	10	5	5	NUM
fcis-6912	120	11	):	):	PUNCT
fcis-6912	120	12	482	482	NUM
fcis-6912	120	13	.	.	PUNCT
fcis-6912	121	1	[	[	X
fcis-6912	121	2	2	2	NUM
fcis-6912	121	3	]	]	X
fcis-6912	121	4	zhang	zhang	PROPN
fcis-6912	121	5	libang	libang	PROPN
fcis-6912	121	6	chinese	chinese	PROPN
fcis-6912	121	7	electronic	electronic	ADJ
fcis-6912	121	8	medical	medical	ADJ
fcis-6912	121	9	record	record	NOUN
fcis-6912	121	10	segmentation	segmentation	NOUN
fcis-6912	121	11	and	and	CCONJ
fcis-6912	121	12	name	name	NOUN
fcis-6912	121	13	entity	entity	NOUN
fcis-6912	121	14	mining	mining	NOUN
fcis-6912	121	15	based	base	VERB
fcis-6912	121	16	on	on	ADP
fcis-6912	121	17	semi	semi	ADV
fcis-6912	121	18	supervised	supervised	ADJ
fcis-6912	121	19	learning	learn	VERB
fcis-6912	121	20	[	[	X
fcis-6912	121	21	d	d	X
fcis-6912	121	22	]	]	X
fcis-6912	121	23	the	the	DET
fcis-6912	121	24	following	follow	VERB
fcis-6912	121	25	is	be	AUX
fcis-6912	121	26	:	:	PUNCT
fcis-6912	121	27	[	[	X
fcis-6912	121	28	master	master	NOUN
fcis-6912	121	29	's	's	PART
fcis-6912	121	30	thesis	thesis	NOUN
fcis-6912	121	31	]	]	X
fcis-6912	121	32	harbin	harbin	PROPN
fcis-6912	121	33	:	:	PUNCT
fcis-6912	121	34	harbin	harbin	PROPN
fcis-6912	121	35	institute	institute	PROPN
fcis-6912	121	36	of	of	ADP
fcis-6912	121	37	technology	technology	PROPN
fcis-6912	121	38	,	,	PUNCT
fcis-6912	121	39	2014	2014	NUM
fcis-6912	122	1	[	[	X
fcis-6912	122	2	3	3	X
fcis-6912	122	3	]	]	X
fcis-6912	122	4	chinchor	chinchor	PROPN
fcis-6912	122	5	n.	n.	PROPN
fcis-6912	122	6	muc-6	muc-6	PUNCT
fcis-6912	122	7	named	name	VERB
fcis-6912	122	8	entity	entity	NOUN
fcis-6912	122	9	task	task	NOUN
fcis-6912	122	10	definition	definition	NOUN
fcis-6912	122	11	(	(	PUNCT
fcis-6912	122	12	version	version	NOUN
fcis-6912	122	13	2.1	2.1	NUM
fcis-6912	122	14	)	)	PUNCT
fcis-6912	123	1	[	[	X
fcis-6912	123	2	c	c	X
fcis-6912	123	3	]	]	PUNCT
fcis-6912	123	4	.	.	PUNCT
fcis-6912	124	1	proceedings	proceeding	NOUN
fcis-6912	124	2	of	of	ADP
fcis-6912	124	3	the	the	DET
fcis-6912	124	4	6th	6th	ADJ
fcis-6912	124	5	conference	conference	NOUN
fcis-6912	124	6	on	on	ADP
fcis-6912	124	7	message	message	NOUN
fcis-6912	124	8	understanding	understanding	NOUN
fcis-6912	124	9	,	,	PUNCT
fcis-6912	124	10	columbia	columbia	PROPN
fcis-6912	124	11	,	,	PUNCT
fcis-6912	124	12	maryland	maryland	PROPN
fcis-6912	124	13	,	,	PUNCT
fcis-6912	124	14	1995	1995	NUM
fcis-6912	124	15	:	:	PUNCT
fcis-6912	124	16	142	142	NUM
fcis-6912	124	17	-	-	SYM
fcis-6912	124	18	194	194	NUM
fcis-6912	124	19	.	.	PUNCT
fcis-6912	125	1	[	[	X
fcis-6912	125	2	4	4	NUM
fcis-6912	125	3	]	]	X
fcis-6912	125	4	ldc	ldc	PROPN
fcis-6912	125	5	corporation	corporation	PROPN
fcis-6912	125	6	.	.	PUNCT
fcis-6912	126	1	entity	entity	NOUN
fcis-6912	126	2	detection	detection	NOUN
fcis-6912	126	3	and	and	CCONJ
fcis-6912	126	4	tracking	tracking	NOUN
fcis-6912	126	5	phase	phase	NOUN
fcis-6912	126	6	1	1	NUM
fcis-6912	126	7	experimental	experimental	ADJ
fcis-6912	126	8	study	study	NOUN
fcis-6912	126	9	task	task	NOUN
fcis-6912	126	10	definition	definition	NOUN
fcis-6912	127	1	[	[	X
fcis-6912	127	2	eb	eb	PROPN
fcis-6912	127	3	/	/	SYM
fcis-6912	127	4	ol	ol	PROPN
fcis-6912	127	5	]	]	PUNCT
fcis-6912	127	6	.	.	PUNCT
fcis-6912	128	1	https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/edtphase1-v2.2.pdf	https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/edtphase1-v2.2.pdf	PROPN
fcis-6912	128	2	,	,	PUNCT
fcis-6912	128	3	2017	2017	NUM
fcis-6912	128	4	-	-	SYM
fcis-6912	128	5	03	03	NUM
fcis-6912	128	6	-	-	PUNCT
fcis-6912	128	7	10	10	NUM
fcis-6912	128	8	.	.	PUNCT
fcis-6912	129	1	[	[	X
fcis-6912	129	2	5	5	NUM
fcis-6912	129	3	]	]	PUNCT
fcis-6912	129	4	sang	sang	NOUN
fcis-6912	129	5	,	,	PUNCT
fcis-6912	129	6	eric	eric	PROPN
fcis-6912	129	7	,	,	PUNCT
fcis-6912	129	8	meldfen	meldfen	PROPN
fcis-6912	129	9	.	.	PUNCT
fcis-6912	130	1	conll-2003	conll-2003	NOUN
fcis-6912	130	2	shared	share	VERB
fcis-6912	130	3	task	task	NOUN
fcis-6912	130	4	introduction	introduction	NOUN
fcis-6912	130	5	:	:	PUNCT
fcis-6912	130	6	language	language	NOUN
fcis-6912	130	7	independent	independent	ADJ
fcis-6912	130	8	named	name	VERB
fcis-6912	130	9	entity	entity	NOUN
fcis-6912	130	10	recognition	recognition	NOUN
fcis-6912	131	1	[	[	X
fcis-6912	131	2	c	c	X
fcis-6912	131	3	]	]	PUNCT
fcis-6912	131	4	.	.	PUNCT
fcis-6912	132	1	calculate	calculate	VERB
fcis-6912	132	2	natural	natural	ADJ
fcis-6912	132	3	language	language	NOUN
fcis-6912	132	4	learning	learning	NOUN
fcis-6912	132	5	progress	progress	NOUN
fcis-6912	132	6	.	.	PUNCT
fcis-6912	133	1	association	association	NOUN
fcis-6912	133	2	for	for	ADP
fcis-6912	133	3	computational	computational	ADJ
fcis-6912	133	4	linguistics	linguistic	NOUN
fcis-6912	133	5	,	,	PUNCT
fcis-6912	133	6	2003:142	2003:142	NUM
fcis-6912	133	7	-	-	SYM
fcis-6912	133	8	147	147	NUM
fcis-6912	133	9	.	.	PUNCT
fcis-6912	134	1	[	[	X
fcis-6912	134	2	6	6	NUM
fcis-6912	134	3	]	]	X
fcis-6912	134	4	doddington	doddington	PROPN
fcis-6912	134	5	g.	g.	PROPN
fcis-6912	134	6	tasks	task	NOUN
fcis-6912	134	7	,	,	PUNCT
fcis-6912	134	8	data	datum	NOUN
fcis-6912	134	9	,	,	PUNCT
fcis-6912	134	10	and	and	CCONJ
fcis-6912	134	11	evaluation	evaluation	NOUN
fcis-6912	134	12	of	of	ADP
fcis-6912	134	13	automated	automate	VERB
fcis-6912	134	14	content	content	NOUN
fcis-6912	134	15	extraction	extraction	NOUN
fcis-6912	134	16	(	(	PUNCT
fcis-6912	134	17	ace	ace	NOUN
fcis-6912	134	18	)	)	PUNCT
fcis-6912	134	19	programs	program	NOUN
fcis-6912	135	1	[	[	X
fcis-6912	135	2	j	j	X
fcis-6912	135	3	]	]	X
fcis-6912	135	4	.	.	PUNCT
fcis-6912	136	1	proc	proc	PROPN
fcis-6912	136	2	lrec	lrec	PROPN
fcis-6912	136	3	,	,	PUNCT
fcis-6912	136	4	2004	2004	NUM
fcis-6912	136	5	,	,	PUNCT
fcis-6912	136	6	34(3	34(3	NUM
fcis-6912	136	7	):	):	PUNCT
fcis-6912	136	8	123	123	NUM
fcis-6912	136	9	-	-	SYM
fcis-6912	136	10	153	153	NUM
fcis-6912	136	11	.	.	PUNCT
fcis-6912	137	1	[	[	X
fcis-6912	137	2	7	7	X
fcis-6912	137	3	]	]	X
fcis-6912	137	4	liu	liu	PROPN
fcis-6912	137	5	liu	liu	PROPN
fcis-6912	137	6	,	,	PUNCT
fcis-6912	137	7	wang	wang	PROPN
fcis-6912	137	8	dongbo	dongbo	PROPN
fcis-6912	137	9	overview	overview	NOUN
fcis-6912	137	10	of	of	ADP
fcis-6912	137	11	named	name	VERB
fcis-6912	137	12	entity	entity	NOUN
fcis-6912	137	13	recognition	recognition	NOUN
fcis-6912	137	14	research	research	NOUN
fcis-6912	138	1	[	[	X
fcis-6912	138	2	j	j	X
fcis-6912	138	3	]	]	X
fcis-6912	138	4	.	.	PUNCT
fcis-6912	139	1	journal	journal	PROPN
fcis-6912	139	2	of	of	ADP
fcis-6912	139	3	information	information	NOUN
fcis-6912	139	4	technology	technology	NOUN
fcis-6912	139	5	,	,	PUNCT
fcis-6912	139	6	2018	2018	NUM
fcis-6912	139	7	,	,	PUNCT
fcis-6912	139	8	37	37	NUM
fcis-6912	139	9	(	(	PUNCT
fcis-6912	139	10	03	03	NUM
fcis-6912	139	11	):	):	PUNCT
fcis-6912	139	12	329	329	NUM
fcis-6912	139	13	-	-	SYM
fcis-6912	139	14	340	340	NUM
fcis-6912	140	1	[	[	NOUN
fcis-6912	140	2	8	8	NUM
fcis-6912	140	3	]	]	PUNCT
fcis-6912	140	4	silviu	silviu	NOUN
fcis-6912	140	5	cucerzan	cucerzan	NOUN
fcis-6912	140	6	,	,	PUNCT
fcis-6912	140	7	david	david	PROPN
fcis-6912	140	8	yarowsky	yarowsky	PROPN
fcis-6912	140	9	.	.	PUNCT
fcis-6912	141	1	language	language	NOUN
fcis-6912	141	2	independent	independent	ADJ
fcis-6912	141	3	named	name	VERB
fcis-6912	141	4	entity	entity	NOUN
fcis-6912	141	5	recognition	recognition	NOUN
fcis-6912	141	6	combining	combine	VERB
fcis-6912	141	7	morphological	morphological	ADJ
fcis-6912	141	8	and	and	CCONJ
fcis-6912	141	9	contextual	contextual	ADJ
fcis-6912	141	10	evidence	evidence	NOUN
fcis-6912	141	11	[	[	X
fcis-6912	141	12	c	c	X
fcis-6912	141	13	]	]	PUNCT
fcis-6912	141	14	.	.	PUNCT
fcis-6912	142	1	proceedings	proceeding	NOUN
fcis-6912	142	2	of	of	ADP
fcis-6912	142	3	the	the	DET
fcis-6912	142	4	1999	1999	NUM
fcis-6912	142	5	sigdat	sigdat	NOUN
fcis-6912	142	6	emnlp	emnlp	NOUN
fcis-6912	142	7	and	and	CCONJ
fcis-6912	142	8	vlc	vlc	PROPN
fcis-6912	142	9	joint	joint	ADJ
fcis-6912	142	10	meeting	meeting	NOUN
fcis-6912	142	11	,	,	PUNCT
fcis-6912	142	12	1999:90	1999:90	NUM
fcis-6912	142	13	-	-	SYM
fcis-6912	142	14	99	99	NUM
fcis-6912	142	15	.	.	PUNCT
fcis-6912	143	1	[	[	X
fcis-6912	143	2	9	9	NUM
fcis-6912	143	3	]	]	SYM
fcis-6912	143	4	li	li	NOUN
fcis-6912	143	5	d	d	PROPN
fcis-6912	143	6	,	,	PUNCT
fcis-6912	143	7	savova	savova	VERB
fcis-6912	143	8	g	g	ADP
fcis-6912	143	9	,	,	PUNCT
fcis-6912	143	10	kipper	kipper	NOUN
fcis-6912	143	11	-	-	PUNCT
fcis-6912	143	12	schuler	schuler	PROPN
fcis-6912	143	13	k.	k.	PROPN
fcis-6912	143	14	conditional	conditional	ADJ
fcis-6912	143	15	random	random	ADJ
fcis-6912	143	16	fields	field	NOUN
fcis-6912	143	17	and	and	CCONJ
fcis-6912	143	18	support	support	VERB
fcis-6912	143	19	vector	vector	NOUN
fcis-6912	143	20	machines	machine	NOUN
fcis-6912	143	21	for	for	ADP
fcis-6912	143	22	recognition	recognition	NOUN
fcis-6912	143	23	of	of	ADP
fcis-6912	143	24	named	name	VERB
fcis-6912	143	25	obstacle	obstacle	NOUN
fcis-6912	143	26	entities	entity	NOUN
fcis-6912	143	27	in	in	ADP
fcis-6912	143	28	clinical	clinical	ADJ
fcis-6912	143	29	text	text	NOUN
fcis-6912	144	1	[	[	X
fcis-6912	144	2	c	c	X
fcis-6912	144	3	]	]	PUNCT
fcis-6912	144	4	.	.	PUNCT
fcis-6912	145	1	proceedings	proceeding	NOUN
fcis-6912	145	2	of	of	ADP
fcis-6912	145	3	the	the	DET
fcis-6912	145	4	2008	2008	NUM
fcis-6912	145	5	symposium	symposium	NOUN
fcis-6912	145	6	on	on	ADP
fcis-6912	145	7	current	current	ADJ
fcis-6912	145	8	trends	trend	NOUN
fcis-6912	145	9	in	in	ADP
fcis-6912	145	10	biomedical	biomedical	ADJ
fcis-6912	145	11	natural	natural	ADJ
fcis-6912	145	12	language	language	NOUN
fcis-6912	145	13	processing	processing	NOUN
fcis-6912	145	14	.	.	PUNCT
fcis-6912	146	1	2008:94	2008:94	NUM
fcis-6912	146	2	-	-	SYM
fcis-6912	146	3	95	95	NUM
fcis-6912	146	4	.	.	PUNCT
fcis-6912	147	1	[	[	X
fcis-6912	147	2	10	10	NUM
fcis-6912	147	3	]	]	X
fcis-6912	147	4	jiang	jiang	PROPN
fcis-6912	147	5	m	m	PROPN
fcis-6912	147	6	,	,	PUNCT
fcis-6912	147	7	chen	chen	PROPN
fcis-6912	147	8	y	y	PROPN
fcis-6912	147	9	,	,	PUNCT
fcis-6912	147	10	liu	liu	PROPN
fcis-6912	147	11	m.	m.	PROPN
fcis-6912	147	12	research	research	PROPN
fcis-6912	147	13	on	on	ADP
fcis-6912	147	14	machine	machine	NOUN
fcis-6912	147	15	learning	learning	NOUN
fcis-6912	147	16	based	base	VERB
fcis-6912	147	17	methods	method	NOUN
fcis-6912	147	18	to	to	PART
fcis-6912	147	19	extract	extract	VERB
fcis-6912	147	20	clinical	clinical	ADJ
fcis-6912	147	21	entities	entity	NOUN
fcis-6912	147	22	and	and	CCONJ
fcis-6912	147	23	their	their	PRON
fcis-6912	147	24	assertions	assertion	NOUN
fcis-6912	147	25	from	from	ADP
fcis-6912	147	26	a	a	DET
fcis-6912	147	27	large	large	ADJ
fcis-6912	147	28	number	number	NOUN
fcis-6912	147	29	of	of	ADP
fcis-6912	147	30	abstracts	abstract	NOUN
fcis-6912	148	1	[	[	X
fcis-6912	148	2	j	j	X
fcis-6912	148	3	]	]	X
fcis-6912	148	4	.	.	PUNCT
fcis-6912	149	1	journal	journal	PROPN
fcis-6912	149	2	of	of	ADP
fcis-6912	149	3	the	the	DET
fcis-6912	149	4	american	american	PROPN
fcis-6912	149	5	medical	medical	PROPN
fcis-6912	149	6	informatics	informatics	PROPN
fcis-6912	149	7	association	association	PROPN
fcis-6912	149	8	,	,	PUNCT
fcis-6912	149	9	2011	2011	NUM
fcis-6912	149	10	,	,	PUNCT
fcis-6912	149	11	18	18	NUM
fcis-6912	149	12	(	(	PUNCT
fcis-6912	149	13	5	5	NUM
fcis-6912	149	14	):	):	PUNCT
fcis-6912	149	15	601	601	NUM
fcis-6912	149	16	-	-	NUM
fcis-6912	149	17	606	606	NUM
fcis-6912	149	18	.	.	PUNCT
fcis-6912	150	1	[	[	X
fcis-6912	150	2	11	11	NUM
fcis-6912	150	3	]	]	X
fcis-6912	150	4	li	li	PROPN
fcis-6912	150	5	l	l	PROPN
fcis-6912	150	6	,	,	PUNCT
fcis-6912	150	7	mao	mao	PROPN
fcis-6912	150	8	t	t	PROPN
fcis-6912	150	9	,	,	PUNCT
fcis-6912	150	10	huang	huang	PROPN
fcis-6912	150	11	d	d	PROPN
fcis-6912	150	12	,	,	PUNCT
fcis-6912	150	13	et	et	PROPN
fcis-6912	150	14	al	al	PROPN
fcis-6912	150	15	.	.	PUNCT
fcis-6912	151	1	a	a	DET
fcis-6912	151	2	hybrid	hybrid	ADJ
fcis-6912	151	3	model	model	NOUN
fcis-6912	151	4	for	for	ADP
fcis-6912	151	5	chinese	chinese	PROPN
fcis-6912	151	6	named	name	VERB
fcis-6912	151	7	entity	entity	NOUN
fcis-6912	151	8	recognition	recognition	NOUN
fcis-6912	152	1	[	[	X
fcis-6912	152	2	c	c	X
fcis-6912	152	3	]	]	PUNCT
fcis-6912	152	4	.	.	PUNCT
fcis-6912	153	1	proceedings	proceeding	NOUN
fcis-6912	153	2	of	of	ADP
fcis-6912	153	3	the	the	DET
fcis-6912	153	4	5th	5th	ADJ
fcis-6912	153	5	sighan	sighan	ADJ
fcis-6912	153	6	symposium	symposium	NOUN
fcis-6912	153	7	on	on	ADP
fcis-6912	153	8	chinese	chinese	ADJ
fcis-6912	153	9	language	language	NOUN
fcis-6912	153	10	processing	processing	NOUN
fcis-6912	153	11	.	.	PUNCT
fcis-6912	154	1	strauss	strauss	PROPN
fcis-6912	154	2	burg	burg	PROPN
fcis-6912	154	3	:	:	PUNCT
fcis-6912	154	4	computational	computational	ADJ
fcis-6912	154	5	linguistics	linguistics	PROPN
fcis-6912	154	6	association	association	PROPN
fcis-6912	154	7	,	,	PUNCT
fcis-6912	154	8	2006:7278	2006:7278	NUM
fcis-6912	154	9	.	.	PUNCT
fcis-6912	155	1	[	[	X
fcis-6912	155	2	12	12	NUM
fcis-6912	155	3	]	]	X
fcis-6912	155	4	cucerzan	cucerzan	NOUN
fcis-6912	155	5	s.	s.	PROPN
fcis-6912	155	6	large	large	ADJ
fcis-6912	155	7	scale	scale	NOUN
fcis-6912	155	8	named	name	VERB
fcis-6912	155	9	entity	entity	NOUN
fcis-6912	155	10	disambiguation	disambiguation	NOUN
fcis-6912	155	11	based	base	VERB
fcis-6912	155	12	on	on	ADP
fcis-6912	155	13	wikipedia	wikipedia	PROPN
fcis-6912	155	14	data	data	PROPN
fcis-6912	156	1	[	[	X
fcis-6912	156	2	c	c	X
fcis-6912	156	3	]	]	PUNCT
fcis-6912	156	4	.	.	PUNCT
fcis-6912	157	1	proceedings	proceeding	NOUN
fcis-6912	157	2	on	on	ADP
fcis-6912	157	3	empirical	empirical	ADJ
fcis-6912	157	4	methods	method	NOUN
fcis-6912	157	5	in	in	ADP
fcis-6912	157	6	natural	natural	ADJ
fcis-6912	157	7	language	language	NOUN
fcis-6912	157	8	processing	processing	NOUN
fcis-6912	157	9	.	.	PUNCT
fcis-6912	158	1	czech	czech	PROPN
fcis-6912	158	2	republic	republic	PROPN
fcis-6912	158	3	:	:	PUNCT
fcis-6912	158	4	prague	prague	PROPN
fcis-6912	158	5	,	,	PUNCT
fcis-6912	158	6	2007:708	2007:708	NOUN
fcis-6912	158	7	-	-	PUNCT
fcis-6912	158	8	716	716	NUM
fcis-6912	158	9	.	.	PUNCT
fcis-6912	159	1	[	[	X
fcis-6912	159	2	13	13	NUM
fcis-6912	159	3	]	]	X
fcis-6912	159	4	ronan	ronan	PROPN
fcis-6912	159	5	collabert	collabert	PROPN
fcis-6912	159	6	,	,	PUNCT
fcis-6912	159	7	jason	jason	PROPN
fcis-6912	159	8	weston	weston	PROPN
fcis-6912	159	9	,	,	PUNCT
fcis-6912	159	10	l	l	PROPN
fcis-6912	159	11	é	é	PROPN
fcis-6912	159	12	on	on	ADP
fcis-6912	159	13	bottou	bottou	NOUN
fcis-6912	159	14	,	,	PUNCT
fcis-6912	159	15	et	et	PROPN
fcis-6912	159	16	al	al	PROPN
fcis-6912	159	17	.	.	PROPN
fcis-6912	159	18	natural	natural	ADJ
fcis-6912	159	19	language	language	NOUN
fcis-6912	159	20	processing	processing	NOUN
fcis-6912	159	21	starts	start	NOUN
fcis-6912	159	22	(	(	PUNCT
fcis-6912	159	23	almost	almost	ADV
fcis-6912	159	24	)	)	PUNCT
fcis-6912	159	25	from	from	ADP
fcis-6912	159	26	scratch	scratch	NOUN
fcis-6912	159	27	[	[	X
fcis-6912	159	28	j	j	X
fcis-6912	159	29	]	]	X
fcis-6912	159	30	.	.	PUNCT
fcis-6912	159	31	journal	journal	PROPN
fcis-6912	159	32	of	of	ADP
fcis-6912	159	33	machine	machine	NOUN
fcis-6912	159	34	learning	learn	VERB
fcis-6912	159	35	research	research	NOUN
fcis-6912	159	36	,	,	PUNCT
fcis-6912	159	37	2011	2011	NUM
fcis-6912	159	38	,	,	PUNCT
fcis-6912	159	39	12	12	NUM
fcis-6912	159	40	(	(	PUNCT
fcis-6912	159	41	76	76	NUM
fcis-6912	159	42	):	):	PUNCT
fcis-6912	159	43	2493	2493	NUM
fcis-6912	159	44	-	-	SYM
fcis-6912	159	45	2537	2537	NUM
fcis-6912	159	46	.	.	PUNCT
fcis-6912	160	1	[	[	X
fcis-6912	160	2	14	14	NUM
fcis-6912	160	3	]	]	X
fcis-6912	160	4	sahu	sahu	PROPN
fcis-6912	160	5	s	s	PROPN
fcis-6912	160	6	k	k	PROPN
fcis-6912	160	7	,	,	PUNCT
fcis-6912	160	8	anand	anand	PROPN
fcis-6912	160	9	a.	a.	PROPN
fcis-6912	160	10	recursive	recursive	PROPN
fcis-6912	160	11	neural	neural	ADJ
fcis-6912	160	12	network	network	NOUN
fcis-6912	160	13	model	model	NOUN
fcis-6912	160	14	for	for	ADP
fcis-6912	160	15	disease	disease	NOUN
fcis-6912	160	16	name	name	NOUN
fcis-6912	160	17	recognition	recognition	NOUN
fcis-6912	160	18	using	use	VERB
fcis-6912	160	19	domain	domain	NOUN
fcis-6912	160	20	invariant	invariant	ADJ
fcis-6912	160	21	features	feature	NOUN
fcis-6912	160	22	[	[	X
fcis-6912	160	23	c	c	X
fcis-6912	160	24	]	]	PUNCT
fcis-6912	160	25	.	.	PUNCT
fcis-6912	161	1	proceedings	proceeding	NOUN
fcis-6912	161	2	of	of	ADP
fcis-6912	161	3	the	the	DET
fcis-6912	161	4	54th	54th	ADJ
fcis-6912	161	5	annual	annual	ADJ
fcis-6912	161	6	conference	conference	NOUN
fcis-6912	161	7	of	of	ADP
fcis-6912	161	8	the	the	DET
fcis-6912	161	9	computational	computational	ADJ
fcis-6912	161	10	linguistics	linguistics	PROPN
fcis-6912	161	11	association	association	PROPN
fcis-6912	161	12	.	.	PUNCT
fcis-6912	162	1	strauss	strauss	PROPN
fcis-6912	162	2	burg	burg	PROPN
fcis-6912	162	3	:	:	PUNCT
fcis-6912	162	4	computational	computational	ADJ
fcis-6912	162	5	linguistics	linguistics	PROPN
fcis-6912	162	6	association	association	PROPN
fcis-6912	162	7	,	,	PUNCT
fcis-6912	162	8	2016	2016	NUM
fcis-6912	162	9	:	:	PUNCT
fcis-6912	162	10	2216	2216	NUM
fcis-6912	162	11	-	-	SYM
fcis-6912	162	12	2225	2225	NUM
fcis-6912	162	13	.	.	PUNCT
fcis-6912	163	1	[	[	X
fcis-6912	163	2	15	15	NUM
fcis-6912	163	3	]	]	X
fcis-6912	163	4	vaswani	vaswani	NOUN
fcis-6912	163	5	a	a	PRON
fcis-6912	163	6	,	,	PUNCT
fcis-6912	163	7	shazeer	shazeer	NOUN
fcis-6912	163	8	n	n	SYM
fcis-6912	163	9	,	,	PUNCT
fcis-6912	163	10	parmar	parmar	PROPN
fcis-6912	163	11	n	n	CCONJ
fcis-6912	163	12	,	,	PUNCT
fcis-6912	163	13	etc	etc	X
fcis-6912	163	14	.	.	X
fcis-6912	163	15	attention	attention	NOUN
fcis-6912	163	16	is	be	AUX
fcis-6912	163	17	what	what	PRON
fcis-6912	163	18	you	you	PRON
fcis-6912	163	19	need	need	VERB
fcis-6912	163	20	[	[	X
fcis-6912	163	21	j	j	X
fcis-6912	163	22	]	]	X
fcis-6912	163	23	.	.	PUNCT
fcis-6912	164	1	ar	ar	PROPN
fcis-6912	164	2	xiv	xiv	PROPN
fcis-6912	164	3	,	,	PUNCT
fcis-6912	164	4	2017	2017	NUM
fcis-6912	164	5	,	,	PUNCT
fcis-6912	164	6	(	(	PUNCT
fcis-6912	164	7	1706	1706	NUM
fcis-6912	164	8	):	):	PUNCT
fcis-6912	164	9	37	37	NUM
fcis-6912	164	10	-	-	SYM
fcis-6912	164	11	62	62	NUM
fcis-6912	164	12	.	.	PUNCT
fcis-6912	165	1	[	[	X
fcis-6912	165	2	16	16	NUM
fcis-6912	165	3	]	]	X
fcis-6912	165	4	yan	yan	PROPN
fcis-6912	165	5	h	h	PROPN
fcis-6912	165	6	,	,	PUNCT
fcis-6912	165	7	deng	deng	PROPN
fcis-6912	165	8	b	b	PROPN
fcis-6912	165	9	,	,	PUNCT
fcis-6912	165	10	li	li	PROPN
fcis-6912	165	11	x	x	PROPN
fcis-6912	165	12	,	,	PUNCT
fcis-6912	165	13	et	et	PROPN
fcis-6912	165	14	al	al	PROPN
fcis-6912	165	15	.	.	PROPN
fcis-6912	165	16	tener	tener	PROPN
fcis-6912	165	17	:	:	PUNCT
fcis-6912	165	18	adaptive	adaptive	ADJ
fcis-6912	165	19	converter	converter	NOUN
fcis-6912	165	20	encoder	encoder	NOUN
fcis-6912	165	21	for	for	ADP
fcis-6912	165	22	name	name	NOUN
fcis-6912	165	23	entity	entity	NOUN
fcis-6912	165	24	recognition	recognition	NOUN
fcis-6912	166	1	[	[	X
fcis-6912	166	2	j	j	X
fcis-6912	166	3	]	]	X
fcis-6912	166	4	.	.	PUNCT
fcis-6912	167	1	ar	ar	PROPN
fcis-6912	167	2	xiv	xiv	PROPN
fcis-6912	167	3	,	,	PUNCT
fcis-6912	167	4	2019	2019	NUM
fcis-6912	167	5	,	,	PUNCT
fcis-6912	167	6	(	(	PUNCT
fcis-6912	167	7	1911	1911	NUM
fcis-6912	167	8	):	):	PUNCT
fcis-6912	167	9	44	44	NUM
fcis-6912	167	10	-	-	SYM
fcis-6912	167	11	74	74	NUM
fcis-6912	167	12	.	.	PUNCT
fcis-6912	168	1	[	[	X
fcis-6912	168	2	17	17	NUM
fcis-6912	168	3	]	]	X
fcis-6912	168	4	bharadwaj	bharadwaj	PROPN
fcis-6912	168	5	a	a	PROPN
fcis-6912	168	6	,	,	PUNCT
fcis-6912	168	7	mortensen	mortensen	PROPN
fcis-6912	168	8	d	d	PROPN
fcis-6912	168	9	,	,	PUNCT
fcis-6912	168	10	dyer	dyer	NOUN
fcis-6912	168	11	c	c	NOUN
fcis-6912	168	12	,	,	PUNCT
fcis-6912	168	13	etc	etc	X
fcis-6912	168	14	.	.	X
fcis-6912	168	15	speech	speech	NOUN
fcis-6912	168	16	perceptual	perceptual	ADJ
fcis-6912	168	17	neural	neural	ADJ
fcis-6912	168	18	model	model	NOUN
fcis-6912	168	19	for	for	ADP
fcis-6912	168	20	named	name	VERB
fcis-6912	168	21	entity	entity	NOUN
fcis-6912	168	22	recognition	recognition	NOUN
fcis-6912	168	23	in	in	ADP
fcis-6912	168	24	low	low	ADJ
fcis-6912	168	25	resource	resource	NOUN
fcis-6912	168	26	transfer	transfer	NOUN
fcis-6912	168	27	environments	environment	NOUN
fcis-6912	169	1	[	[	X
fcis-6912	169	2	c	c	X
fcis-6912	169	3	]	]	PUNCT
fcis-6912	169	4	.	.	PUNCT
fcis-6912	170	1	proceedings	proceeding	NOUN
fcis-6912	170	2	of	of	ADP
fcis-6912	170	3	the	the	DET
fcis-6912	170	4	2016	2016	NUM
fcis-6912	170	5	conference	conference	NOUN
fcis-6912	170	6	on	on	ADP
fcis-6912	170	7	empirical	empirical	ADJ
fcis-6912	170	8	methods	method	NOUN
fcis-6912	170	9	of	of	ADP
fcis-6912	170	10	natural	natural	ADJ
fcis-6912	170	11	language	language	NOUN
fcis-6912	170	12	processing	processing	NOUN
fcis-6912	170	13	.	.	PUNCT
fcis-6912	171	1	strauss	strauss	PROPN
fcis-6912	171	2	burg	burg	PROPN
fcis-6912	171	3	:	:	PUNCT
fcis-6912	171	4	computational	computational	ADJ
fcis-6912	171	5	linguistics	linguistics	PROPN
fcis-6912	171	6	association	association	PROPN
fcis-6912	171	7	,	,	PUNCT
fcis-6912	171	8	2016:1462	2016:1462	NOUN
fcis-6912	171	9	-	-	NOUN
fcis-6912	171	10	1472	1472	NUM
fcis-6912	171	11	.	.	PUNCT
fcis-6912	172	1	[	[	X
fcis-6912	172	2	18	18	NUM
fcis-6912	172	3	]	]	X
fcis-6912	172	4	devlin	devlin	PROPN
fcis-6912	172	5	j	j	PROPN
fcis-6912	172	6	,	,	PUNCT
fcis-6912	172	7	chang	chang	PROPN
fcis-6912	172	8	m	m	PROPN
fcis-6912	172	9	w	w	PROPN
fcis-6912	172	10	,	,	PUNCT
fcis-6912	172	11	lee	lee	PROPN
fcis-6912	172	12	k	k	PROPN
fcis-6912	172	13	,	,	PUNCT
fcis-6912	172	14	et	et	PROPN
fcis-6912	172	15	al	al	PROPN
fcis-6912	172	16	.	.	PUNCT
fcis-6912	172	17	bert	bert	PROPN
fcis-6912	172	18	:	:	PUNCT
fcis-6912	172	19	pre	pre	VERB
fcis-6912	172	20	training	training	NOUN
fcis-6912	172	21	of	of	ADP
fcis-6912	172	22	deep	deep	ADJ
fcis-6912	172	23	bidirectional	bidirectional	ADJ
fcis-6912	172	24	converters	converter	NOUN
fcis-6912	172	25	for	for	ADP
fcis-6912	172	26	language	language	NOUN
fcis-6912	172	27	understanding	understanding	NOUN
fcis-6912	172	28	[	[	X
fcis-6912	172	29	j	j	X
fcis-6912	172	30	]	]	X
fcis-6912	172	31	.	.	PUNCT
fcis-6912	173	1	ar	ar	PROPN
fcis-6912	173	2	xiv	xiv	PROPN
fcis-6912	173	3	,	,	PUNCT
fcis-6912	173	4	2018	2018	NUM
fcis-6912	173	5	,	,	PUNCT
fcis-6912	173	6	(	(	PUNCT
fcis-6912	173	7	1810	1810	NUM
fcis-6912	173	8	):	):	PUNCT
fcis-6912	173	9	48	48	NUM
fcis-6912	173	10	-	-	SYM
fcis-6912	173	11	50	50	NUM
fcis-6912	173	12	.	.	PUNCT
fcis-6912	174	1	[	[	X
fcis-6912	174	2	19	19	NUM
fcis-6912	174	3	]	]	X
fcis-6912	174	4	liu	liu	PROPN
fcis-6912	174	5	y	y	PROPN
fcis-6912	174	6	h	h	PROPN
fcis-6912	174	7	,	,	PUNCT
fcis-6912	174	8	ott	ott	PROPN
fcis-6912	174	9	m	m	PROPN
fcis-6912	174	10	,	,	PUNCT
fcis-6912	174	11	goyal	goyal	PROPN
fcis-6912	174	12	n	n	CCONJ
fcis-6912	174	13	,	,	PUNCT
fcis-6912	174	14	et	et	PROPN
fcis-6912	174	15	al	al	PROPN
fcis-6912	174	16	.	.	PUNCT
fcis-6912	175	1	roberta	roberta	PROPN
fcis-6912	175	2	:	:	PUNCT
fcis-6912	175	3	a	a	DET
fcis-6912	175	4	robust	robust	ADJ
fcis-6912	175	5	optimization	optimization	NOUN
fcis-6912	175	6	bert	bert	NOUN
fcis-6912	175	7	pre	pre	ADJ
fcis-6912	175	8	training	training	NOUN
fcis-6912	175	9	method	method	NOUN
fcis-6912	175	10	[	[	X
fcis-6912	175	11	eb	eb	PROPN
fcis-6912	175	12	/	/	SYM
fcis-6912	175	13	ol	ol	PROPN
fcis-6912	175	14	]	]	PUNCT
fcis-6912	175	15	.	.	PUNCT
fcis-6912	176	1	https://arxiv.org/abs/1907.11692	https://arxiv.org/abs/1907.11692	PROPN
fcis-6912	176	2	,	,	PUNCT
fcis-6912	176	3	2019	2019	NUM
fcis-6912	176	4	-	-	SYM
fcis-6912	176	5	07	07	NUM
fcis-6912	176	6	-	-	PUNCT
fcis-6912	176	7	26	26	NUM
fcis-6912	176	8	.	.	PUNCT
fcis-6912	177	1	[	[	X
fcis-6912	177	2	20	20	NUM
fcis-6912	177	3	]	]	X
fcis-6912	177	4	lan	lan	PROPN
fcis-6912	177	5	z	z	PROPN
fcis-6912	177	6	,	,	PUNCT
fcis-6912	177	7	chen	chen	PROPN
fcis-6912	177	8	m	m	PROPN
fcis-6912	177	9	,	,	PUNCT
fcis-6912	177	10	goodman	goodman	PROPN
fcis-6912	177	11	s	s	PROPN
fcis-6912	177	12	,	,	PUNCT
fcis-6912	178	1	et	et	PROPN
fcis-6912	178	2	al	al	PROPN
fcis-6912	178	3	.	.	PROPN
fcis-6912	178	4	albert	albert	PROPN
fcis-6912	178	5	:	:	PUNCT
fcis-6912	178	6	simplified	simplified	ADJ
fcis-6912	178	7	bert	bert	PROPN
fcis-6912	179	1	[	[	X
fcis-6912	179	2	eb	eb	PROPN
fcis-6912	179	3	/	/	SYM
fcis-6912	179	4	ol	ol	PROPN
fcis-6912	179	5	]	]	X
fcis-6912	179	6	for	for	ADP
fcis-6912	179	7	language	language	NOUN
fcis-6912	179	8	representation	representation	NOUN
fcis-6912	179	9	self	self	NOUN
fcis-6912	179	10	supervised	supervise	VERB
fcis-6912	179	11	learning	learning	NOUN
fcis-6912	179	12	.	.	PUNCT
fcis-6912	180	1	https://openreview.net/pdf?id=h1ea7aetvs	https://openreview.net/pdf?id=h1ea7aetvs	NOUN
fcis-6912	180	2	,	,	PUNCT
fcis-6912	180	3	2019	2019	NUM
fcis-6912	180	4	-	-	SYM
fcis-6912	180	5	07	07	NUM
fcis-6912	180	6	-	-	PUNCT
fcis-6912	180	7	26	26	NUM
fcis-6912	180	8	.	.	PUNCT
fcis-6912	181	1	[	[	X
fcis-6912	181	2	21	21	NUM
fcis-6912	181	3	]	]	X
fcis-6912	181	4	wang	wang	PROPN
fcis-6912	181	5	naiyu	naiyu	PROPN
fcis-6912	181	6	,	,	PUNCT
fcis-6912	181	7	ye	ye	PROPN
fcis-6912	181	8	yuxin	yuxin	PROPN
fcis-6912	181	9	,	,	PUNCT
fcis-6912	181	10	liu	liu	PROPN
fcis-6912	181	11	lu	lu	PROPN
fcis-6912	181	12	,	,	PUNCT
fcis-6912	181	13	feng	feng	PROPN
fcis-6912	181	14	lizhou	lizhou	PROPN
fcis-6912	181	15	,	,	PUNCT
fcis-6912	181	16	bao	bao	PROPN
fcis-6912	181	17	tie	tie	PROPN
fcis-6912	181	18	,	,	PUNCT
fcis-6912	181	19	peng	peng	PROPN
fcis-6912	181	20	tao	tao	PROPN
fcis-6912	181	21	.	.	PUNCT
fcis-6912	182	1	research	research	NOUN
fcis-6912	182	2	on	on	ADP
fcis-6912	182	3	language	language	NOUN
fcis-6912	182	4	models	model	NOUN
fcis-6912	182	5	based	base	VERB
fcis-6912	182	6	on	on	ADP
fcis-6912	182	7	deep	deep	ADJ
fcis-6912	182	8	learning	learning	NOUN
fcis-6912	182	9	progress	progress	NOUN
fcis-6912	183	1	[	[	X
fcis-6912	183	2	j	j	X
fcis-6912	183	3	]	]	X
fcis-6912	183	4	.	.	PUNCT
fcis-6912	184	1	journal	journal	PROPN
fcis-6912	184	2	of	of	ADP
fcis-6912	184	3	software	software	NOUN
fcis-6912	184	4	,	,	PUNCT
fcis-6912	184	5	2021	2021	NUM
fcis-6912	184	6	,	,	PUNCT
fcis-6912	184	7	32	32	NUM
fcis-6912	184	8	(	(	PUNCT
fcis-6912	184	9	04	04	NUM
fcis-6912	184	10	):	):	PUNCT
fcis-6912	184	11	1082	1082	NUM
fcis-6912	184	12	-	-	SYM
fcis-6912	184	13	1115	1115	NUM
fcis-6912	184	14	[	[	X
fcis-6912	184	15	22	22	NUM
fcis-6912	184	16	]	]	PUNCT
fcis-6912	184	17	dongl	dongl	PROPN
fcis-6912	184	18	,	,	PUNCT
fcis-6912	184	19	yangn	yangn	PROPN
fcis-6912	184	20	,	,	PUNCT
fcis-6912	184	21	wangw	wangw	NOUN
fcis-6912	184	22	,	,	PUNCT
fcis-6912	184	23	et	et	PROPN
fcis-6912	184	24	al	al	PROPN
fcis-6912	184	25	.	.	PUNCT
fcis-6912	185	1	a	a	DET
fcis-6912	185	2	unified	unified	ADJ
fcis-6912	185	3	language	language	NOUN
fcis-6912	185	4	model	model	NOUN
fcis-6912	185	5	for	for	ADP
fcis-6912	185	6	pre	pre	X
fcis-6912	185	7	training	training	NOUN
fcis-6912	185	8	and	and	CCONJ
fcis-6912	185	9	generation	generation	NOUN
fcis-6912	185	10	of	of	ADP
fcis-6912	185	11	natural	natural	ADJ
fcis-6912	185	12	languages	language	NOUN
fcis-6912	185	13	[	[	X
fcis-6912	185	14	eb	eb	X
fcis-6912	185	15	/	/	SYM
fcis-6912	185	16	ol	ol	PROPN
fcis-6912	185	17	]	]	PUNCT
fcis-6912	185	18	.	.	PUNCT
fcis-6912	186	1	https://arxiv.org/abs/1905.03197	https://arxiv.org/abs/1905.03197	PROPN
fcis-6912	186	2	,	,	PUNCT
fcis-6912	186	3	2019	2019	NUM
fcis-6912	186	4	-	-	SYM
fcis-6912	186	5	0	0	NUM
fcis-6912	186	6	5	5	NUM
fcis-6912	186	7	-	-	SYM
fcis-6912	186	8	08	08	NUM
fcis-6912	186	9	.	.	PUNCT
fcis-6912	187	1	[	[	X
fcis-6912	187	2	23	23	NUM
fcis-6912	187	3	]	]	PUNCT
fcis-6912	187	4	song	song	NOUN
fcis-6912	187	5	k	k	PROPN
fcis-6912	187	6	,	,	PUNCT
fcis-6912	187	7	tan	tan	PROPN
fcis-6912	187	8	x	x	SYM
fcis-6912	187	9	,	,	PUNCT
fcis-6912	187	10	qin	qin	PROPN
fcis-6912	187	11	t	t	PROPN
fcis-6912	187	12	,	,	PUNCT
fcis-6912	187	13	et	et	PROPN
fcis-6912	187	14	al	al	PROPN
fcis-6912	187	15	.	.	PROPN
fcis-6912	187	16	mass	mass	PROPN
fcis-6912	187	17	:	:	PUNCT
fcis-6912	187	18	language	language	NOUN
fcis-6912	187	19	generated	generate	VERB
fcis-6912	187	20	masking	mask	VERB
fcis-6912	187	21	sequence	sequence	NOUN
fcis-6912	187	22	to	to	ADP
fcis-6912	187	23	sequence	sequence	NOUN
fcis-6912	187	24	pre	pre	ADJ
fcis-6912	187	25	training	training	NOUN
fcis-6912	187	26	[	[	X
fcis-6912	187	27	eb	eb	PROPN
fcis-6912	187	28	/	/	SYM
fcis-6912	187	29	ol	ol	PROPN
fcis-6912	187	30	]	]	PUNCT
fcis-6912	187	31	.	.	PUNCT
fcis-6912	188	1	https://arxiv.org/pdf/1905.02450.pdf	https://arxiv.org/pdf/1905.02450.pdf	PROPN
fcis-6912	188	2	,	,	PUNCT
fcis-6912	188	3	2019	2019	NUM
fcis-6912	188	4	-	-	SYM
fcis-6912	188	5	06	06	NUM
fcis-6912	188	6	-	-	SYM
fcis-6912	188	7	21	21	NUM
fcis-6912	188	8	.	.	PUNCT
fcis-6912	189	1	[	[	X
fcis-6912	189	2	24	24	NUM
fcis-6912	189	3	]	]	PUNCT
fcis-6912	189	4	tsai	tsai	PROPN
fcis-6912	189	5	h	h	PROPN
fcis-6912	189	6	,	,	PUNCT
fcis-6912	189	7	riesa	riesa	PROPN
fcis-6912	189	8	j	j	PROPN
fcis-6912	189	9	,	,	PUNCT
fcis-6912	189	10	johnson	johnson	PROPN
fcis-6912	189	11	m	m	PROPN
fcis-6912	189	12	,	,	PUNCT
fcis-6912	189	13	et	et	PROPN
fcis-6912	189	14	al	al	PROPN
fcis-6912	189	15	.	.	PUNCT
fcis-6912	190	1	a	a	DET
fcis-6912	190	2	small	small	ADJ
fcis-6912	190	3	practical	practical	ADJ
fcis-6912	190	4	bert	bert	NOUN
fcis-6912	190	5	model	model	NOUN
fcis-6912	190	6	for	for	ADP
fcis-6912	190	7	sequence	sequence	NOUN
fcis-6912	190	8	tagging	tag	VERB
fcis-6912	190	9	[	[	X
fcis-6912	190	10	c	c	X
fcis-6912	190	11	]	]	PUNCT
fcis-6912	190	12	.	.	PUNCT
fcis-6912	191	1	procedure	procedure	NOUN
fcis-6912	191	2	.	.	PUNCT
fcis-6912	192	1	2019	2019	NUM
fcis-6912	192	2	conference	conference	NOUN
fcis-6912	192	3	on	on	ADP
fcis-6912	192	4	natural	natural	ADJ
fcis-6912	192	5	language	language	NOUN
fcis-6912	192	6	processing	processing	NOUN
fcis-6912	192	7	experiences	experience	NOUN
fcis-6912	192	8	and	and	CCONJ
fcis-6912	192	9	methods	method	NOUN
fcis-6912	192	10	and	and	CCONJ
fcis-6912	192	11	the	the	DET
fcis-6912	192	12	ninth	ninth	ADJ
fcis-6912	192	13	international	international	ADJ
fcis-6912	192	14	joint	joint	ADJ
fcis-6912	192	15	conference	conference	NOUN
fcis-6912	192	16	on	on	ADP
fcis-6912	192	17	natural	natural	ADJ
fcis-6912	192	18	language	language	NOUN
fcis-6912	192	19	processing	processing	NOUN
fcis-6912	192	20	(	(	PUNCT
fcis-6912	192	21	emnlp	emnlp	NOUN
fcis-6912	192	22	-	-	PUNCT
fcis-6912	192	23	ijcnlp	ijcnlp	NOUN
fcis-6912	192	24	)	)	PUNCT
fcis-6912	192	25	,	,	PUNCT
fcis-6912	192	26	2019	2019	NUM
fcis-6912	192	27	:	:	SYM
fcis-6912	192	28	3623	3623	NUM
fcis-6912	192	29	–	–	PUNCT
fcis-6912	192	30	3627	3627	NUM
fcis-6912	192	31	.	.	PUNCT
fcis-6912	193	1	[	[	X
fcis-6912	193	2	25	25	NUM
fcis-6912	193	3	]	]	X
fcis-6912	193	4	lu	lu	PROPN
fcis-6912	193	5	w	w	PROPN
fcis-6912	193	6	,	,	PUNCT
fcis-6912	193	7	dan	dan	PROPN
fcis-6912	193	8	r.	r.	PROPN
fcis-6912	193	9	joint	joint	PROPN
fcis-6912	193	10	reference	reference	NOUN
fcis-6912	193	11	extraction	extraction	NOUN
fcis-6912	193	12	and	and	CCONJ
fcis-6912	193	13	classification	classification	NOUN
fcis-6912	193	14	using	use	VERB
fcis-6912	193	15	reference	reference	NOUN
fcis-6912	193	16	hypergraphs	hypergraph	NOUN
fcis-6912	194	1	[	[	X
fcis-6912	194	2	c	c	X
fcis-6912	194	3	]	]	PUNCT
fcis-6912	194	4	.	.	PUNCT
fcis-6912	195	1	proceedings	proceeding	NOUN
fcis-6912	195	2	of	of	ADP
fcis-6912	195	3	the	the	DET
fcis-6912	195	4	2015	2015	NUM
fcis-6912	195	5	conference	conference	NOUN
fcis-6912	195	6	on	on	ADP
fcis-6912	195	7	empirical	empirical	ADJ
fcis-6912	195	8	methods	method	NOUN
fcis-6912	195	9	of	of	ADP
fcis-6912	195	10	natural	natural	ADJ
fcis-6912	195	11	language	language	NOUN
fcis-6912	195	12	processing	processing	NOUN
fcis-6912	195	13	,	,	PUNCT
fcis-6912	195	14	2015:857	2015:857	NUM
fcis-6912	195	15	-	-	SYM
fcis-6912	195	16	867	867	NUM
fcis-6912	195	17	.	.	PUNCT
fcis-6912	196	1	[	[	X
fcis-6912	196	2	26	26	NUM
fcis-6912	196	3	]	]	PUNCT
fcis-6912	196	4	muis	muis	PROPN
fcis-6912	196	5	a	a	DET
fcis-6912	196	6	o	o	NOUN
fcis-6912	196	7	,	,	PUNCT
fcis-6912	196	8	lu	lu	PROPN
fcis-6912	196	9	w.	w.	PROPN
fcis-6912	196	10	mark	mark	PROPN
fcis-6912	196	11	gaps	gap	NOUN
fcis-6912	196	12	between	between	ADP
fcis-6912	196	13	words	word	NOUN
fcis-6912	196	14	:	:	PUNCT
fcis-6912	196	15	use	use	VERB
fcis-6912	196	16	reference	reference	NOUN
fcis-6912	196	17	separators	separator	NOUN
fcis-6912	196	18	to	to	PART
fcis-6912	196	19	identify	identify	VERB
fcis-6912	196	20	overlapping	overlap	VERB
fcis-6912	196	21	references	reference	NOUN
fcis-6912	196	22	[	[	X
fcis-6912	196	23	c	c	X
fcis-6912	196	24	]	]	PUNCT
fcis-6912	196	25	.	.	PUNCT
fcis-6912	197	1	proceedings	proceeding	NOUN
fcis-6912	197	2	of	of	ADP
fcis-6912	197	3	the	the	DET
fcis-6912	197	4	2017	2017	NUM
fcis-6912	197	5	conference	conference	NOUN
fcis-6912	197	6	on	on	ADP
fcis-6912	197	7	empirical	empirical	ADJ
fcis-6912	197	8	methods	method	NOUN
fcis-6912	197	9	of	of	ADP
fcis-6912	197	10	natural	natural	ADJ
fcis-6912	197	11	language	language	NOUN
fcis-6912	197	12	processing	processing	NOUN
fcis-6912	197	13	,	,	PUNCT
fcis-6912	197	14	2017:2608	2017:2608	PROPN
fcis-6912	197	15	-	-	SYM
fcis-6912	197	16	2618	2618	NUM
fcis-6912	197	17	.	.	PUNCT
fcis-6912	198	1	[	[	X
fcis-6912	198	2	27	27	NUM
fcis-6912	198	3	]	]	X
fcis-6912	198	4	ju	ju	PROPN
fcis-6912	198	5	m	m	PROPN
fcis-6912	198	6	,	,	PUNCT
fcis-6912	198	7	miwa	miwa	PROPN
fcis-6912	198	8	m	m	PROPN
fcis-6912	198	9	,	,	PUNCT
fcis-6912	198	10	ananiadou	ananiadou	NOUN
fcis-6912	198	11	s.	s.	PROPN
fcis-6912	198	12	neural	neural	PROPN
fcis-6912	198	13	hierarchical	hierarchical	ADJ
fcis-6912	198	14	model	model	NOUN
fcis-6912	198	15	for	for	ADP
fcis-6912	198	16	nested	nested	ADJ
fcis-6912	198	17	named	name	VERB
fcis-6912	198	18	entity	entity	NOUN
fcis-6912	198	19	recognition	recognition	NOUN
fcis-6912	199	1	[	[	X
fcis-6912	199	2	c	c	X
fcis-6912	199	3	]	]	PUNCT
fcis-6912	199	4	.	.	PUNCT
fcis-6912	200	1	proceedings	proceeding	NOUN
fcis-6912	200	2	of	of	ADP
fcis-6912	200	3	the	the	DET
fcis-6912	200	4	2018	2018	NUM
fcis-6912	200	5	north	north	NOUN
fcis-6912	200	6	american	american	ADJ
fcis-6912	200	7	branch	branch	NOUN
fcis-6912	200	8	of	of	ADP
fcis-6912	200	9	the	the	DET
fcis-6912	200	10	computational	computational	ADJ
fcis-6912	200	11	linguistics	linguistics	PROPN
fcis-6912	200	12	association	association	PROPN
fcis-6912	200	13	:	:	PUNCT
fcis-6912	200	14	human	human	ADJ
fcis-6912	200	15	language	language	NOUN
fcis-6912	200	16	technology	technology	NOUN
fcis-6912	200	17	,	,	PUNCT
fcis-6912	200	18	volume	volume	NOUN
fcis-6912	200	19	1	1	NUM
fcis-6912	200	20	(	(	PUNCT
fcis-6912	200	21	long	long	ADJ
fcis-6912	200	22	paper	paper	NOUN
fcis-6912	200	23	)	)	PUNCT
fcis-6912	200	24	,	,	PUNCT
fcis-6912	200	25	2018	2018	NUM
fcis-6912	200	26	:	:	PUNCT
fcis-6912	200	27	1446	1446	NUM
fcis-6912	200	28	-	-	SYM
fcis-6912	200	29	1459	1459	NUM
fcis-6912	200	30	.	.	PUNCT
fcis-6912	201	1	[	[	X
fcis-6912	201	2	28	28	NUM
fcis-6912	201	3	]	]	X
fcis-6912	201	4	jue	jue	PROPN
fcis-6912	201	5	w	w	PROPN
fcis-6912	201	6	,	,	PUNCT
fcis-6912	201	7	shou	shou	PROPN
fcis-6912	201	8	l	l	PROPN
fcis-6912	201	9	,	,	PUNCT
fcis-6912	201	10	chen	chen	PROPN
fcis-6912	201	11	k	k	PROPN
fcis-6912	201	12	,	,	PUNCT
fcis-6912	201	13	et	et	PROPN
fcis-6912	201	14	al	al	PROPN
fcis-6912	201	15	.	.	PUNCT
fcis-6912	201	16	pyramid	pyramid	PROPN
fcis-6912	201	17	:	:	PUNCT
fcis-6912	201	18	a	a	DET
fcis-6912	201	19	layered	layered	ADJ
fcis-6912	201	20	model	model	NOUN
fcis-6912	201	21	for	for	ADP
fcis-6912	201	22	nested	nested	ADJ
fcis-6912	201	23	named	name	VERB
fcis-6912	201	24	entity	entity	NOUN
fcis-6912	201	25	recognition[c	recognition[c	PROPN
fcis-6912	201	26	]	]	PUNCT
fcis-6912	201	27	.	.	PUNCT
fcis-6912	202	1	proceedings	proceeding	NOUN
fcis-6912	202	2	of	of	ADP
fcis-6912	202	3	the	the	DET
fcis-6912	202	4	58th	58th	ADJ
fcis-6912	202	5	annual	annual	ADJ
fcis-6912	202	6	meeting	meeting	NOUN
fcis-6912	202	7	of	of	ADP
fcis-6912	202	8	the	the	DET
fcis-6912	202	9	association	association	NOUN
fcis-6912	202	10	for	for	ADP
fcis-6912	202	11	computational	computational	ADJ
fcis-6912	202	12	linguistics	linguistic	NOUN
fcis-6912	202	13	,	,	PUNCT
fcis-6912	202	14	2020:5918	2020:5918	NUM
fcis-6912	202	15	-	-	SYM
fcis-6912	202	16	5928	5928	NUM
fcis-6912	202	17	.	.	PUNCT
fcis-6912	203	1	[	[	X
fcis-6912	203	2	29	29	NUM
fcis-6912	203	3	]	]	X
fcis-6912	203	4	li	li	PROPN
fcis-6912	203	5	zhoujun	zhoujun	PROPN
fcis-6912	203	6	,	,	PUNCT
fcis-6912	203	7	fan	fan	PROPN
fcis-6912	203	8	yu	yu	PROPN
fcis-6912	203	9	,	,	PUNCT
fcis-6912	203	10	wu	wu	PROPN
fcis-6912	203	11	xianjie	xianjie	PROPN
fcis-6912	203	12	.	.	PUNCT
fcis-6912	204	1	overview	overview	NOUN
fcis-6912	204	2	of	of	ADP
fcis-6912	204	3	research	research	NOUN
fcis-6912	204	4	on	on	ADP
fcis-6912	204	5	pretraining	pretraine	VERB
fcis-6912	204	6	technology	technology	NOUN
fcis-6912	204	7	for	for	ADP
fcis-6912	204	8	natural	natural	ADJ
fcis-6912	204	9	language	language	NOUN
fcis-6912	204	10	processing	processing	NOUN
fcis-6912	204	11	[	[	X
fcis-6912	204	12	j	j	X
fcis-6912	204	13	]	]	X
fcis-6912	204	14	.	.	PUNCT
fcis-6912	205	1	computer	computer	NOUN
fcis-6912	205	2	science	science	NOUN
fcis-6912	205	3	,	,	PUNCT
fcis-6912	205	4	2020,47	2020,47	NUM
fcis-6912	205	5	(	(	PUNCT
fcis-6912	205	6	03	03	NUM
fcis-6912	205	7	):	):	PUNCT
fcis-6912	205	8	162	162	NUM
fcis-6912	205	9	-	-	SYM
fcis-6912	205	10	173	173	NUM
fcis-6912	206	1	[	[	X
fcis-6912	206	2	30	30	NUM
fcis-6912	206	3	]	]	X
fcis-6912	206	4	sun	sun	PROPN
fcis-6912	206	5	y	y	PROPN
fcis-6912	206	6	,	,	PUNCT
fcis-6912	206	7	wang	wang	PROPN
fcis-6912	206	8	s	s	PROPN
fcis-6912	206	9	,	,	PUNCT
fcis-6912	206	10	li	li	PROPN
fcis-6912	206	11	y	y	PROPN
fcis-6912	206	12	,	,	PUNCT
fcis-6912	206	13	et	et	PROPN
fcis-6912	206	14	al	al	PROPN
fcis-6912	206	15	.	.	PROPN
fcis-6912	206	16	ernie	ernie	PROPN
fcis-6912	206	17	2.0	2.0	NUM
fcis-6912	206	18	:	:	PUNCT
fcis-6912	206	19	a	a	DET
fcis-6912	206	20	continual	continual	ADJ
fcis-6912	206	21	pretrain⁃ing	pretrain⁃ing	NOUN
fcis-6912	206	22	framework	framework	NOUN
fcis-6912	206	23	for	for	ADP
fcis-6912	206	24	language	language	NOUN
fcis-6912	206	25	understanding[j	understanding[j	NOUN
fcis-6912	206	26	]	]	PUNCT
fcis-6912	206	27	.	.	PUNCT
fcis-6912	207	1	proceedings	proceeding	NOUN
fcis-6912	207	2	of	of	ADP
fcis-6912	207	3	theaaai	theaaai	NOUN
fcis-6912	207	4	conference	conference	NOUN
fcis-6912	207	5	on	on	ADP
fcis-6912	207	6	artifical	artifical	ADJ
fcis-6912	207	7	intelligence	intelligence	NOUN
fcis-6912	207	8	,	,	PUNCT
fcis-6912	207	9	2020	2020	NUM
fcis-6912	207	10	,	,	PUNCT
fcis-6912	207	11	34(5	34(5	NUM
fcis-6912	207	12	):	):	PUNCT
fcis-6912	207	13	8968	8968	NUM
fcis-6912	207	14	-	-	SYM
fcis-6912	207	15	8975	8975	NUM
fcis-6912	207	16	.	.	PUNCT
fcis-6912	208	1	[	[	X
fcis-6912	208	2	31	31	NUM
fcis-6912	208	3	]	]	X
fcis-6912	208	4	luo	luo	PROPN
fcis-6912	208	5	xiao	xiao	PROPN
fcis-6912	208	6	.	.	PROPN
fcis-6912	208	7	overview	overview	NOUN
fcis-6912	208	8	of	of	ADP
fcis-6912	208	9	natural	natural	ADJ
fcis-6912	208	10	language	language	NOUN
fcis-6912	208	11	processing	processing	NOUN
fcis-6912	208	12	research	research	NOUN
fcis-6912	208	13	based	base	VERB
fcis-6912	208	14	on	on	ADP
fcis-6912	208	15	deep	deep	ADJ
fcis-6912	208	16	learning	learning	NOUN
fcis-6912	208	17	[	[	X
fcis-6912	208	18	j	j	X
fcis-6912	208	19	]	]	X
fcis-6912	208	20	.	.	PUNCT
fcis-6912	209	1	intelligent	intelligent	ADJ
fcis-6912	209	2	computers	computer	NOUN
fcis-6912	209	3	and	and	CCONJ
fcis-6912	209	4	applications	application	NOUN
fcis-6912	209	5	,	,	PUNCT
fcis-6912	209	6	2020	2020	NUM
fcis-6912	209	7	,	,	PUNCT
fcis-6912	209	8	10	10	NUM
fcis-6912	209	9	(	(	PUNCT
fcis-6912	209	10	4	4	NUM
fcis-6912	209	11	):	):	PUNCT
fcis-6912	209	12	133	133	NUM
fcis-6912	209	13	-	-	SYM
fcis-6912	209	14	137	137	NUM
fcis-6912	209	15	.	.	PUNCT
fcis-6912	210	1	luo	luo	PROPN
fcis-6912	210	2	xiao	xiao	PROPN
fcis-6912	210	3	a	a	DET
fcis-6912	210	4	survey	survey	NOUN
fcis-6912	210	5	of	of	ADP
fcis-6912	210	6	natural	natural	ADJ
fcis-6912	210	7	language	language	NOUN
fcis-6912	210	8	processing	processing	NOUN
fcis-6912	210	9	based	base	VERB
fcis-6912	210	10	on	on	ADP
fcis-6912	210	11	deep	deep	ADJ
fcis-6912	210	12	learning[j	learning[j	NOUN
fcis-6912	210	13	]	]	PUNCT
fcis-6912	210	14	.	.	PUNCT
fcis-6912	211	1	intelligent	intelligent	ADJ
fcis-6912	211	2	computer	computer	NOUN
fcis-6912	211	3	and	and	CCONJ
fcis-6912	211	4	application	application	NOUN
fcis-6912	211	5	,	,	PUNCT
fcis-6912	211	6	2020	2020	NUM
fcis-6912	211	7	,	,	PUNCT
fcis-6912	211	8	10(4	10(4	NUM
fcis-6912	211	9	):	):	PUNCT
fcis-6912	211	10	133	133	NUM
fcis-6912	211	11	-	-	SYM
fcis-6912	211	12	137	137	NUM
fcis-6912	211	13	.	.	PUNCT