id	sid	tid	token	lemma	pos
fcis-9479	1	1	frontiers	frontier	NOUN
fcis-9479	1	2	in	in	ADP
fcis-9479	1	3	computing	computing	NOUN
fcis-9479	1	4	and	and	CCONJ
fcis-9479	1	5	intelligent	intelligent	ADJ
fcis-9479	1	6	systems	system	NOUN
fcis-9479	1	7	issn	issn	VERB
fcis-9479	1	8	:	:	PUNCT
fcis-9479	1	9	2832	2832	NUM
fcis-9479	1	10	-	-	SYM
fcis-9479	1	11	6024	6024	NUM
fcis-9479	1	12	|	|	NOUN
fcis-9479	1	13	vol	vol	NOUN
fcis-9479	1	14	.	.	PROPN
fcis-9479	2	1	4	4	NUM
fcis-9479	2	2	,	,	PUNCT
fcis-9479	2	3	no	no	INTJ
fcis-9479	2	4	.	.	NOUN
fcis-9479	2	5	1	1	NUM
fcis-9479	2	6	,	,	PUNCT
fcis-9479	2	7	2023	2023	NUM
fcis-9479	2	8	105	105	NUM
fcis-9479	2	9	fake	fake	ADJ
fcis-9479	2	10	news	news	NOUN
fcis-9479	2	11	detection	detection	NOUN
fcis-9479	2	12	based	base	VERB
fcis-9479	2	13	on	on	ADP
fcis-9479	2	14	deep	deep	ADJ
fcis-9479	2	15	learning	learning	NOUN
fcis-9479	2	16	rong	rong	PROPN
fcis-9479	3	1	wang	wang	PROPN
fcis-9479	3	2	*	*	PUNCT
fcis-9479	3	3	school	school	NOUN
fcis-9479	3	4	of	of	ADP
fcis-9479	3	5	information	information	NOUN
fcis-9479	3	6	,	,	PUNCT
fcis-9479	3	7	yunnan	yunnan	PROPN
fcis-9479	3	8	university	university	PROPN
fcis-9479	3	9	of	of	ADP
fcis-9479	3	10	finance	finance	NOUN
fcis-9479	3	11	and	and	CCONJ
fcis-9479	3	12	economics	economic	NOUN
fcis-9479	3	13	,	,	PUNCT
fcis-9479	3	14	kunming	kunming	NOUN
fcis-9479	3	15	650221	650221	NUM
fcis-9479	3	16	,	,	PUNCT
fcis-9479	3	17	china	china	PROPN
fcis-9479	3	18	*	*	PUNCT
fcis-9479	3	19	corresponding	correspond	VERB
fcis-9479	3	20	author	author	NOUN
fcis-9479	3	21	email	email	NOUN
fcis-9479	3	22	:	:	PUNCT
fcis-9479	3	23	wr594998@163.com	wr594998@163.com	X
fcis-9479	3	24	abstract	abstract	ADJ
fcis-9479	3	25	:	:	PUNCT
fcis-9479	3	26	the	the	DET
fcis-9479	3	27	rapid	rapid	ADJ
fcis-9479	3	28	popularization	popularization	NOUN
fcis-9479	3	29	of	of	ADP
fcis-9479	3	30	the	the	DET
fcis-9479	3	31	internet	internet	NOUN
fcis-9479	3	32	has	have	AUX
fcis-9479	3	33	broken	break	VERB
fcis-9479	3	34	the	the	DET
fcis-9479	3	35	professional	professional	ADJ
fcis-9479	3	36	threshold	threshold	NOUN
fcis-9479	3	37	of	of	ADP
fcis-9479	3	38	information	information	NOUN
fcis-9479	3	39	dissemination	dissemination	NOUN
fcis-9479	3	40	,	,	PUNCT
fcis-9479	3	41	enabling	enable	VERB
fcis-9479	3	42	more	more	ADJ
fcis-9479	3	43	and	and	CCONJ
fcis-9479	3	44	more	more	ADJ
fcis-9479	3	45	people	people	NOUN
fcis-9479	3	46	to	to	PART
fcis-9479	3	47	easily	easily	ADV
fcis-9479	3	48	obtain	obtain	VERB
fcis-9479	3	49	information	information	NOUN
fcis-9479	3	50	,	,	PUNCT
fcis-9479	3	51	share	share	VERB
fcis-9479	3	52	and	and	CCONJ
fcis-9479	3	53	express	express	VERB
fcis-9479	3	54	views	view	NOUN
fcis-9479	3	55	through	through	ADP
fcis-9479	3	56	social	social	ADJ
fcis-9479	3	57	media	medium	NOUN
fcis-9479	3	58	,	,	PUNCT
fcis-9479	3	59	which	which	PRON
fcis-9479	3	60	has	have	AUX
fcis-9479	3	61	greatly	greatly	ADV
fcis-9479	3	62	enriched	enrich	VERB
fcis-9479	3	63	people	people	NOUN
fcis-9479	3	64	's	's	PART
fcis-9479	3	65	daily	daily	ADJ
fcis-9479	3	66	life	life	NOUN
fcis-9479	3	67	.	.	PUNCT
fcis-9479	4	1	however	however	ADV
fcis-9479	4	2	,	,	PUNCT
fcis-9479	4	3	due	due	ADP
fcis-9479	4	4	to	to	ADP
fcis-9479	4	5	the	the	DET
fcis-9479	4	6	huge	huge	ADJ
fcis-9479	4	7	number	number	NOUN
fcis-9479	4	8	of	of	ADP
fcis-9479	4	9	users	user	NOUN
fcis-9479	4	10	of	of	ADP
fcis-9479	4	11	social	social	ADJ
fcis-9479	4	12	media	medium	NOUN
fcis-9479	4	13	,	,	PUNCT
fcis-9479	4	14	false	false	ADJ
fcis-9479	4	15	news	news	NOUN
fcis-9479	4	16	fabricated	fabricate	VERB
fcis-9479	4	17	for	for	ADP
fcis-9479	4	18	various	various	ADJ
fcis-9479	4	19	purposes	purpose	NOUN
fcis-9479	4	20	is	be	AUX
fcis-9479	4	21	emerging	emerge	VERB
fcis-9479	4	22	in	in	ADP
fcis-9479	4	23	endlessly	endlessly	ADV
fcis-9479	4	24	.	.	PUNCT
fcis-9479	5	1	moreover	moreover	ADV
fcis-9479	5	2	,	,	PUNCT
fcis-9479	5	3	with	with	ADP
fcis-9479	5	4	the	the	DET
fcis-9479	5	5	progress	progress	NOUN
fcis-9479	5	6	of	of	ADP
fcis-9479	5	7	technology	technology	NOUN
fcis-9479	5	8	,	,	PUNCT
fcis-9479	5	9	false	false	ADJ
fcis-9479	5	10	news	news	NOUN
fcis-9479	5	11	is	be	AUX
fcis-9479	5	12	no	no	ADV
fcis-9479	5	13	longer	long	ADV
fcis-9479	5	14	simply	simply	ADV
fcis-9479	5	15	spread	spread	VERB
fcis-9479	5	16	in	in	ADP
fcis-9479	5	17	the	the	DET
fcis-9479	5	18	form	form	NOUN
fcis-9479	5	19	of	of	ADP
fcis-9479	5	20	text	text	NOUN
fcis-9479	5	21	,	,	PUNCT
fcis-9479	5	22	but	but	CCONJ
fcis-9479	5	23	more	more	ADJ
fcis-9479	5	24	spread	spread	VERB
fcis-9479	5	25	through	through	ADP
fcis-9479	5	26	the	the	DET
fcis-9479	5	27	combination	combination	NOUN
fcis-9479	5	28	of	of	ADP
fcis-9479	5	29	text	text	NOUN
fcis-9479	5	30	,	,	PUNCT
fcis-9479	5	31	pictures	picture	NOUN
fcis-9479	5	32	and	and	CCONJ
fcis-9479	5	33	video	video	NOUN
fcis-9479	5	34	,	,	PUNCT
fcis-9479	5	35	which	which	PRON
fcis-9479	5	36	greatly	greatly	ADV
fcis-9479	5	37	increases	increase	VERB
fcis-9479	5	38	the	the	DET
fcis-9479	5	39	confusion	confusion	NOUN
fcis-9479	5	40	of	of	ADP
fcis-9479	5	41	false	false	ADJ
fcis-9479	5	42	news	news	NOUN
fcis-9479	5	43	.	.	PUNCT
fcis-9479	6	1	the	the	DET
fcis-9479	6	2	experiment	experiment	NOUN
fcis-9479	6	3	in	in	ADP
fcis-9479	6	4	this	this	DET
fcis-9479	6	5	paper	paper	NOUN
fcis-9479	6	6	is	be	AUX
fcis-9479	6	7	based	base	VERB
fcis-9479	6	8	on	on	ADP
fcis-9479	6	9	tensorflow	tensorflow	NOUN
fcis-9479	6	10	to	to	PART
fcis-9479	6	11	detect	detect	VERB
fcis-9479	6	12	false	false	ADJ
fcis-9479	6	13	news	news	NOUN
fcis-9479	6	14	.	.	PUNCT
fcis-9479	7	1	during	during	ADP
fcis-9479	7	2	the	the	DET
fcis-9479	7	3	experiment	experiment	NOUN
fcis-9479	7	4	,	,	PUNCT
fcis-9479	7	5	lr	lr	PROPN
fcis-9479	7	6	was	be	AUX
fcis-9479	7	7	used	use	VERB
fcis-9479	7	8	to	to	PART
fcis-9479	7	9	obtain	obtain	VERB
fcis-9479	7	10	the	the	DET
fcis-9479	7	11	fusion	fusion	NOUN
fcis-9479	7	12	coefficients	coefficient	NOUN
fcis-9479	7	13	of	of	ADP
fcis-9479	7	14	cnn	cnn	PROPN
fcis-9479	7	15	and	and	CCONJ
fcis-9479	7	16	lstm	lstm	NOUN
fcis-9479	7	17	models	model	NOUN
fcis-9479	7	18	,	,	PUNCT
fcis-9479	7	19	that	that	PRON
fcis-9479	7	20	is	be	AUX
fcis-9479	7	21	the	the	DET
fcis-9479	7	22	regression	regression	NOUN
fcis-9479	7	23	coefficient	coefficient	NOUN
fcis-9479	7	24	of	of	ADP
fcis-9479	7	25	lr	lr	NOUN
fcis-9479	7	26	,	,	PUNCT
fcis-9479	7	27	and	and	CCONJ
fcis-9479	7	28	then	then	ADV
fcis-9479	7	29	calculated	calculate	VERB
fcis-9479	7	30	the	the	DET
fcis-9479	7	31	optimal	optimal	ADJ
fcis-9479	7	32	threshold	threshold	NOUN
fcis-9479	7	33	with	with	ADP
fcis-9479	7	34	the	the	DET
fcis-9479	7	35	fused	fuse	VERB
fcis-9479	7	36	model	model	NOUN
fcis-9479	7	37	on	on	ADP
fcis-9479	7	38	the	the	DET
fcis-9479	7	39	verification	verification	NOUN
fcis-9479	7	40	set	set	NOUN
fcis-9479	7	41	.	.	PUNCT
fcis-9479	8	1	in	in	ADP
fcis-9479	8	2	addition	addition	NOUN
fcis-9479	8	3	,	,	PUNCT
fcis-9479	8	4	in	in	ADP
fcis-9479	8	5	terms	term	NOUN
fcis-9479	8	6	of	of	ADP
fcis-9479	8	7	model	model	NOUN
fcis-9479	8	8	selection	selection	NOUN
fcis-9479	8	9	,	,	PUNCT
fcis-9479	8	10	lightgbm	lightgbm	ADJ
fcis-9479	8	11	and	and	CCONJ
fcis-9479	8	12	xgboost	xgboost	PRON
fcis-9479	8	13	were	be	AUX
fcis-9479	8	14	selected	select	VERB
fcis-9479	8	15	to	to	PART
fcis-9479	8	16	train	train	VERB
fcis-9479	8	17	the	the	DET
fcis-9479	8	18	model	model	NOUN
fcis-9479	8	19	on	on	ADP
fcis-9479	8	20	the	the	DET
fcis-9479	8	21	training	training	NOUN
fcis-9479	8	22	set	set	VERB
fcis-9479	8	23	for	for	ADP
fcis-9479	8	24	false	false	ADJ
fcis-9479	8	25	news	news	NOUN
fcis-9479	8	26	,	,	PUNCT
fcis-9479	8	27	and	and	CCONJ
fcis-9479	8	28	predicted	predict	VERB
fcis-9479	8	29	the	the	DET
fcis-9479	8	30	news	news	NOUN
fcis-9479	8	31	text	text	NOUN
fcis-9479	8	32	on	on	ADP
fcis-9479	8	33	the	the	DET
fcis-9479	8	34	testing	testing	NOUN
fcis-9479	8	35	set	set	NOUN
fcis-9479	8	36	.	.	PUNCT
fcis-9479	9	1	the	the	DET
fcis-9479	9	2	results	result	NOUN
fcis-9479	9	3	of	of	ADP
fcis-9479	9	4	three	three	NUM
fcis-9479	9	5	experiments	experiment	NOUN
fcis-9479	9	6	show	show	VERB
fcis-9479	9	7	that	that	SCONJ
fcis-9479	9	8	the	the	DET
fcis-9479	9	9	effect	effect	NOUN
fcis-9479	9	10	of	of	ADP
fcis-9479	9	11	using	use	VERB
fcis-9479	9	12	xgboost	xgboost	PROPN
fcis-9479	9	13	model	model	NOUN
fcis-9479	9	14	is	be	AUX
fcis-9479	9	15	the	the	DET
fcis-9479	9	16	best	good	ADJ
fcis-9479	9	17	,	,	PUNCT
fcis-9479	9	18	and	and	CCONJ
fcis-9479	9	19	the	the	DET
fcis-9479	9	20	f1	f1	PROPN
fcis-9479	9	21	score	score	NOUN
fcis-9479	9	22	obtained	obtain	VERB
fcis-9479	9	23	in	in	ADP
fcis-9479	9	24	the	the	DET
fcis-9479	9	25	experiment	experiment	NOUN
fcis-9479	9	26	is	be	AUX
fcis-9479	9	27	the	the	DET
fcis-9479	9	28	highest	high	ADJ
fcis-9479	9	29	.	.	PUNCT
fcis-9479	10	1	keywords	keyword	NOUN
fcis-9479	10	2	:	:	PUNCT
fcis-9479	10	3	false	false	ADJ
fcis-9479	10	4	news	news	NOUN
fcis-9479	10	5	detection	detection	NOUN
fcis-9479	10	6	;	;	PUNCT
fcis-9479	10	7	tensorflow	tensorflow	NOUN
fcis-9479	10	8	;	;	PUNCT
fcis-9479	10	9	cnn	cnn	PROPN
fcis-9479	10	10	and	and	CCONJ
fcis-9479	10	11	lstm	lstm	NOUN
fcis-9479	10	12	fusion	fusion	NOUN
fcis-9479	10	13	;	;	PUNCT
fcis-9479	10	14	lightgbm	lightgbm	ADJ
fcis-9479	10	15	;	;	PUNCT
fcis-9479	10	16	xgboost	xgboost	X
fcis-9479	10	17	.	.	PUNCT
fcis-9479	11	1	1	1	X
fcis-9479	11	2	.	.	X
fcis-9479	11	3	introduction	introduction	NOUN
fcis-9479	11	4	1.1	1.1	NUM
fcis-9479	11	5	.	.	PUNCT
fcis-9479	12	1	research	research	NOUN
fcis-9479	12	2	background	background	NOUN
fcis-9479	12	3	and	and	CCONJ
fcis-9479	12	4	significance	significance	NOUN
fcis-9479	12	5	with	with	ADP
fcis-9479	12	6	the	the	DET
fcis-9479	12	7	rapid	rapid	ADJ
fcis-9479	12	8	popularity	popularity	NOUN
fcis-9479	12	9	of	of	ADP
fcis-9479	12	10	mobile	mobile	ADJ
fcis-9479	12	11	internet	internet	NOUN
fcis-9479	12	12	,	,	PUNCT
fcis-9479	12	13	people	people	NOUN
fcis-9479	12	14	's	's	PART
fcis-9479	12	15	access	access	NOUN
fcis-9479	12	16	to	to	ADP
fcis-9479	12	17	news	news	NOUN
fcis-9479	12	18	has	have	AUX
fcis-9479	12	19	become	become	VERB
fcis-9479	12	20	more	more	ADV
fcis-9479	12	21	convenient	convenient	ADJ
fcis-9479	12	22	.	.	PUNCT
fcis-9479	13	1	in	in	ADP
fcis-9479	13	2	particular	particular	ADJ
fcis-9479	13	3	,	,	PUNCT
fcis-9479	13	4	social	social	ADJ
fcis-9479	13	5	media	medium	NOUN
fcis-9479	13	6	,	,	PUNCT
fcis-9479	13	7	mainly	mainly	ADV
fcis-9479	13	8	represented	represent	VERB
fcis-9479	13	9	by	by	ADP
fcis-9479	13	10	weibo	weibo	NOUN
fcis-9479	13	11	and	and	CCONJ
fcis-9479	13	12	twitter	twitter	NOUN
fcis-9479	13	13	,	,	PUNCT
fcis-9479	13	14	have	have	AUX
fcis-9479	13	15	become	become	VERB
fcis-9479	13	16	an	an	DET
fcis-9479	13	17	important	important	ADJ
fcis-9479	13	18	channel	channel	NOUN
fcis-9479	13	19	for	for	SCONJ
fcis-9479	13	20	most	most	ADJ
fcis-9479	13	21	people	people	NOUN
fcis-9479	13	22	to	to	PART
fcis-9479	13	23	obtain	obtain	VERB
fcis-9479	13	24	daily	daily	ADJ
fcis-9479	13	25	information	information	NOUN
fcis-9479	13	26	.	.	PUNCT
fcis-9479	14	1	nowadays	nowadays	ADV
fcis-9479	14	2	,	,	PUNCT
fcis-9479	14	3	more	more	ADJ
fcis-9479	14	4	and	and	CCONJ
fcis-9479	14	5	more	more	ADJ
fcis-9479	14	6	people	people	NOUN
fcis-9479	14	7	are	be	AUX
fcis-9479	14	8	gaining	gain	VERB
fcis-9479	14	9	knowledge	knowledge	NOUN
fcis-9479	14	10	,	,	PUNCT
fcis-9479	14	11	sharing	share	VERB
fcis-9479	14	12	information	information	NOUN
fcis-9479	14	13	,	,	PUNCT
fcis-9479	14	14	expressing	express	VERB
fcis-9479	14	15	opinions	opinion	NOUN
fcis-9479	14	16	and	and	CCONJ
fcis-9479	14	17	exchanging	exchange	VERB
fcis-9479	14	18	experiences	experience	NOUN
fcis-9479	14	19	on	on	ADP
fcis-9479	14	20	these	these	DET
fcis-9479	14	21	online	online	ADJ
fcis-9479	14	22	platforms	platform	NOUN
fcis-9479	14	23	every	every	DET
fcis-9479	14	24	day	day	NOUN
fcis-9479	14	25	.	.	PUNCT
fcis-9479	15	1	however	however	ADV
fcis-9479	15	2	,	,	PUNCT
fcis-9479	15	3	to	to	ADP
fcis-9479	15	4	a	a	DET
fcis-9479	15	5	certain	certain	ADJ
fcis-9479	15	6	extent	extent	NOUN
fcis-9479	15	7	,	,	PUNCT
fcis-9479	15	8	the	the	DET
fcis-9479	15	9	rapid	rapid	ADJ
fcis-9479	15	10	development	development	NOUN
fcis-9479	15	11	of	of	ADP
fcis-9479	15	12	social	social	ADJ
fcis-9479	15	13	media	medium	NOUN
fcis-9479	15	14	is	be	AUX
fcis-9479	15	15	a	a	DET
fcis-9479	15	16	double	double	ADJ
fcis-9479	15	17	-	-	PUNCT
fcis-9479	15	18	edged	edge	VERB
fcis-9479	15	19	sword	sword	NOUN
fcis-9479	15	20	.	.	PUNCT
fcis-9479	16	1	on	on	ADP
fcis-9479	16	2	the	the	DET
fcis-9479	16	3	one	one	NUM
fcis-9479	16	4	hand	hand	NOUN
fcis-9479	16	5	,	,	PUNCT
fcis-9479	16	6	it	it	PRON
fcis-9479	16	7	has	have	AUX
fcis-9479	16	8	facilitated	facilitate	VERB
fcis-9479	16	9	the	the	DET
fcis-9479	16	10	advent	advent	NOUN
fcis-9479	16	11	of	of	ADP
fcis-9479	16	12	the	the	DET
fcis-9479	16	13	information	information	NOUN
fcis-9479	16	14	age	age	NOUN
fcis-9479	16	15	,	,	PUNCT
fcis-9479	16	16	where	where	SCONJ
fcis-9479	16	17	all	all	DET
fcis-9479	16	18	kinds	kind	NOUN
fcis-9479	16	19	of	of	ADP
fcis-9479	16	20	information	information	NOUN
fcis-9479	16	21	can	can	AUX
fcis-9479	16	22	be	be	AUX
fcis-9479	16	23	disseminated	disseminate	VERB
fcis-9479	16	24	at	at	ADP
fcis-9479	16	25	a	a	DET
fcis-9479	16	26	very	very	ADV
fcis-9479	16	27	low	low	ADJ
fcis-9479	16	28	cost	cost	NOUN
fcis-9479	16	29	through	through	ADP
fcis-9479	16	30	social	social	ADJ
fcis-9479	16	31	media	medium	NOUN
fcis-9479	16	32	,	,	PUNCT
fcis-9479	16	33	making	make	VERB
fcis-9479	16	34	it	it	PRON
fcis-9479	16	35	possible	possible	ADJ
fcis-9479	16	36	for	for	ADP
fcis-9479	16	37	"	"	PUNCT
fcis-9479	16	38	a	a	DET
fcis-9479	16	39	genius	genius	NOUN
fcis-9479	16	40	to	to	PART
fcis-9479	16	41	know	know	VERB
fcis-9479	16	42	everything	everything	PRON
fcis-9479	16	43	about	about	ADP
fcis-9479	16	44	the	the	DET
fcis-9479	16	45	world	world	NOUN
fcis-9479	16	46	without	without	ADP
fcis-9479	16	47	leaving	leave	VERB
fcis-9479	16	48	home	home	NOUN
fcis-9479	16	49	"	"	PUNCT
fcis-9479	16	50	.	.	PUNCT
fcis-9479	17	1	on	on	ADP
fcis-9479	17	2	the	the	DET
fcis-9479	17	3	other	other	ADJ
fcis-9479	17	4	hand	hand	NOUN
fcis-9479	17	5	,	,	PUNCT
fcis-9479	17	6	the	the	DET
fcis-9479	17	7	internet	internet	NOUN
fcis-9479	17	8	has	have	AUX
fcis-9479	17	9	become	become	VERB
fcis-9479	17	10	a	a	DET
fcis-9479	17	11	breeding	breeding	NOUN
fcis-9479	17	12	ground	ground	NOUN
fcis-9479	18	1	[	[	X
fcis-9479	18	2	1	1	X
fcis-9479	18	3	]	]	PUNCT
fcis-9479	18	4	for	for	ADP
fcis-9479	18	5	false	false	ADJ
fcis-9479	18	6	information	information	NOUN
fcis-9479	18	7	to	to	PART
fcis-9479	18	8	grow	grow	VERB
fcis-9479	18	9	and	and	CCONJ
fcis-9479	18	10	spread	spread	VERB
fcis-9479	18	11	,	,	PUNCT
fcis-9479	18	12	and	and	CCONJ
fcis-9479	18	13	the	the	DET
fcis-9479	18	14	vast	vast	ADJ
fcis-9479	18	15	amount	amount	NOUN
fcis-9479	18	16	of	of	ADP
fcis-9479	18	17	information	information	NOUN
fcis-9479	18	18	is	be	AUX
fcis-9479	18	19	mixed	mix	VERB
fcis-9479	18	20	with	with	ADP
fcis-9479	18	21	false	false	ADJ
fcis-9479	18	22	news	news	NOUN
fcis-9479	18	23	that	that	PRON
fcis-9479	18	24	is	be	AUX
fcis-9479	18	25	deliberately	deliberately	ADV
fcis-9479	18	26	fabricated	fabricate	VERB
fcis-9479	18	27	for	for	ADP
fcis-9479	18	28	various	various	ADJ
fcis-9479	18	29	purposes	purpose	NOUN
fcis-9479	18	30	.	.	PUNCT
fcis-9479	19	1	this	this	DET
fcis-9479	19	2	kind	kind	NOUN
fcis-9479	19	3	of	of	ADP
fcis-9479	19	4	news	news	NOUN
fcis-9479	19	5	is	be	AUX
fcis-9479	19	6	presented	present	VERB
fcis-9479	19	7	in	in	ADP
fcis-9479	19	8	various	various	ADJ
fcis-9479	19	9	forms	form	NOUN
fcis-9479	19	10	,	,	PUNCT
fcis-9479	19	11	often	often	ADV
fcis-9479	19	12	with	with	ADP
fcis-9479	19	13	text	text	NOUN
fcis-9479	19	14	and	and	CCONJ
fcis-9479	19	15	pictures	picture	NOUN
fcis-9479	19	16	,	,	PUNCT
fcis-9479	19	17	which	which	PRON
fcis-9479	19	18	makes	make	VERB
fcis-9479	19	19	it	it	PRON
fcis-9479	19	20	very	very	ADV
fcis-9479	19	21	difficult	difficult	ADJ
fcis-9479	19	22	for	for	SCONJ
fcis-9479	19	23	readers	reader	NOUN
fcis-9479	19	24	to	to	PART
fcis-9479	19	25	get	get	VERB
fcis-9479	19	26	correct	correct	ADJ
fcis-9479	19	27	and	and	CCONJ
fcis-9479	19	28	effective	effective	ADJ
fcis-9479	19	29	information	information	NOUN
fcis-9479	19	30	,	,	PUNCT
fcis-9479	19	31	and	and	CCONJ
fcis-9479	19	32	some	some	PRON
fcis-9479	19	33	of	of	ADP
fcis-9479	19	34	the	the	DET
fcis-9479	19	35	widely	widely	ADV
fcis-9479	19	36	disseminated	disseminate	VERB
fcis-9479	19	37	false	false	ADJ
fcis-9479	19	38	news	news	NOUN
fcis-9479	19	39	has	have	AUX
fcis-9479	19	40	caused	cause	VERB
fcis-9479	19	41	serious	serious	ADJ
fcis-9479	19	42	negative	negative	ADJ
fcis-9479	19	43	impact	impact	NOUN
fcis-9479	19	44	on	on	ADP
fcis-9479	19	45	individuals	individual	NOUN
fcis-9479	19	46	and	and	CCONJ
fcis-9479	19	47	society	society	NOUN
fcis-9479	19	48	.	.	PUNCT
fcis-9479	20	1	thus	thus	ADV
fcis-9479	20	2	,	,	PUNCT
fcis-9479	20	3	it	it	PRON
fcis-9479	20	4	seems	seem	VERB
fcis-9479	20	5	that	that	SCONJ
fcis-9479	20	6	there	there	PRON
fcis-9479	20	7	is	be	VERB
fcis-9479	20	8	a	a	DET
fcis-9479	20	9	great	great	ADJ
fcis-9479	20	10	need	need	NOUN
fcis-9479	20	11	for	for	SCONJ
fcis-9479	20	12	us	we	PRON
fcis-9479	20	13	to	to	PART
fcis-9479	20	14	study	study	VERB
fcis-9479	20	15	the	the	DET
fcis-9479	20	16	fake	fake	ADJ
fcis-9479	20	17	news	news	NOUN
fcis-9479	20	18	detection	detection	NOUN
fcis-9479	20	19	.	.	PUNCT
fcis-9479	21	1	despite	despite	SCONJ
fcis-9479	21	2	some	some	DET
fcis-9479	21	3	recent	recent	ADJ
fcis-9479	21	4	scientific	scientific	ADJ
fcis-9479	21	5	advances	advance	NOUN
fcis-9479	21	6	in	in	ADP
fcis-9479	21	7	false	false	ADJ
fcis-9479	21	8	news	news	NOUN
fcis-9479	21	9	detection	detection	NOUN
fcis-9479	21	10	,	,	PUNCT
fcis-9479	21	11	it	it	PRON
fcis-9479	21	12	is	be	AUX
fcis-9479	21	13	still	still	ADV
fcis-9479	21	14	a	a	DET
fcis-9479	21	15	challenging	challenging	ADJ
fcis-9479	21	16	problem	problem	NOUN
fcis-9479	21	17	due	due	ADP
fcis-9479	21	18	to	to	ADP
fcis-9479	21	19	its	its	PRON
fcis-9479	21	20	complex	complex	ADJ
fcis-9479	21	21	content	content	NOUN
fcis-9479	21	22	,	,	PUNCT
fcis-9479	21	23	wide	wide	ADJ
fcis-9479	21	24	range	range	NOUN
fcis-9479	21	25	of	of	ADP
fcis-9479	21	26	sources	source	NOUN
fcis-9479	21	27	,	,	PUNCT
fcis-9479	21	28	diverse	diverse	ADJ
fcis-9479	21	29	modalities	modality	NOUN
fcis-9479	21	30	,	,	PUNCT
fcis-9479	21	31	and	and	CCONJ
fcis-9479	21	32	the	the	DET
fcis-9479	21	33	high	high	ADJ
fcis-9479	21	34	costs	cost	NOUN
fcis-9479	21	35	involved	involve	VERB
fcis-9479	21	36	in	in	ADP
fcis-9479	21	37	factchecking	factchecke	VERB
fcis-9479	21	38	.	.	PUNCT
fcis-9479	22	1	fake	fake	ADJ
fcis-9479	22	2	news	news	NOUN
fcis-9479	22	3	spreads	spread	VERB
fcis-9479	22	4	rapidly	rapidly	ADV
fcis-9479	22	5	through	through	ADP
fcis-9479	22	6	social	social	ADJ
fcis-9479	22	7	media	medium	NOUN
fcis-9479	22	8	platforms	platform	NOUN
fcis-9479	22	9	,	,	PUNCT
fcis-9479	22	10	which	which	PRON
fcis-9479	22	11	may	may	AUX
fcis-9479	22	12	affect	affect	VERB
fcis-9479	22	13	social	social	ADJ
fcis-9479	22	14	stability	stability	NOUN
fcis-9479	22	15	while	while	SCONJ
fcis-9479	22	16	negatively	negatively	ADV
fcis-9479	22	17	impacting	impact	VERB
fcis-9479	22	18	social	social	ADJ
fcis-9479	22	19	opinion	opinion	NOUN
fcis-9479	22	20	.	.	PUNCT
fcis-9479	23	1	for	for	ADP
fcis-9479	23	2	example	example	NOUN
fcis-9479	23	3	,	,	PUNCT
fcis-9479	23	4	false	false	ADJ
fcis-9479	23	5	news	news	NOUN
fcis-9479	23	6	such	such	ADJ
fcis-9479	23	7	as	as	ADP
fcis-9479	23	8	"	"	PUNCT
fcis-9479	23	9	temporarily	temporarily	ADV
fcis-9479	23	10	exempting	exempt	VERB
fcis-9479	23	11	wuhan	wuhan	PROPN
fcis-9479	23	12	red	red	PROPN
fcis-9479	23	13	cross	cross	PROPN
fcis-9479	23	14	from	from	ADP
fcis-9479	23	15	managing	manage	VERB
fcis-9479	23	16	disaster	disaster	NOUN
fcis-9479	23	17	relief	relief	NOUN
fcis-9479	23	18	supplies	supply	NOUN
fcis-9479	23	19	"	"	PUNCT
fcis-9479	23	20	and	and	CCONJ
fcis-9479	23	21	"	"	PUNCT
fcis-9479	23	22	the	the	DET
fcis-9479	23	23	private	private	ADJ
fcis-9479	23	24	car	car	NOUN
fcis-9479	23	25	was	be	AUX
fcis-9479	23	26	unclaimed	unclaimed	ADJ
fcis-9479	23	27	for	for	ADP
fcis-9479	23	28	several	several	ADJ
fcis-9479	23	29	months	month	NOUN
fcis-9479	23	30	,	,	PUNCT
fcis-9479	23	31	passers	passers	AUX
fcis-9479	23	32	-	-	PUNCT
fcis-9479	23	33	by	by	ADP
fcis-9479	23	34	asked	ask	VERB
fcis-9479	23	35	and	and	CCONJ
fcis-9479	23	36	learned	learn	VERB
fcis-9479	23	37	that	that	SCONJ
fcis-9479	23	38	the	the	DET
fcis-9479	23	39	owner	owner	NOUN
fcis-9479	23	40	died	die	VERB
fcis-9479	23	41	after	after	ADP
fcis-9479	23	42	fighting	fight	VERB
fcis-9479	23	43	the	the	DET
fcis-9479	23	44	epidemic	epidemic	NOUN
fcis-9479	23	45	"	"	PUNCT
fcis-9479	23	46	,	,	PUNCT
fcis-9479	23	47	etc	etc	X
fcis-9479	23	48	.	.	X
fcis-9479	24	1	although	although	SCONJ
fcis-9479	24	2	the	the	DET
fcis-9479	24	3	official	official	NOUN
fcis-9479	24	4	explained	explain	VERB
fcis-9479	24	5	and	and	CCONJ
fcis-9479	24	6	refuted	refute	VERB
fcis-9479	24	7	the	the	DET
fcis-9479	24	8	rumors	rumor	NOUN
fcis-9479	24	9	at	at	ADP
fcis-9479	24	10	the	the	DET
fcis-9479	24	11	first	first	ADJ
fcis-9479	24	12	time	time	NOUN
fcis-9479	24	13	,	,	PUNCT
fcis-9479	24	14	such	such	ADJ
fcis-9479	24	15	rumors	rumor	NOUN
fcis-9479	24	16	still	still	ADV
fcis-9479	24	17	circulated	circulate	VERB
fcis-9479	24	18	widely	widely	ADV
fcis-9479	24	19	on	on	ADP
fcis-9479	24	20	the	the	DET
fcis-9479	24	21	internet	internet	NOUN
fcis-9479	24	22	.	.	PUNCT
fcis-9479	25	1	it	it	PRON
fcis-9479	25	2	has	have	AUX
fcis-9479	25	3	attracted	attract	VERB
fcis-9479	25	4	a	a	DET
fcis-9479	25	5	lot	lot	NOUN
fcis-9479	25	6	of	of	ADP
fcis-9479	25	7	attention	attention	NOUN
fcis-9479	25	8	and	and	CCONJ
fcis-9479	25	9	caused	cause	VERB
fcis-9479	25	10	serious	serious	ADJ
fcis-9479	25	11	negative	negative	ADJ
fcis-9479	25	12	impact	impact	NOUN
fcis-9479	25	13	on	on	ADP
fcis-9479	25	14	the	the	DET
fcis-9479	25	15	parties	party	NOUN
fcis-9479	25	16	concerned	concern	VERB
fcis-9479	25	17	.	.	PUNCT
fcis-9479	26	1	at	at	ADP
fcis-9479	26	2	present	present	ADJ
fcis-9479	26	3	,	,	PUNCT
fcis-9479	26	4	most	most	ADJ
fcis-9479	26	5	of	of	ADP
fcis-9479	26	6	the	the	DET
fcis-9479	26	7	news	news	NOUN
fcis-9479	26	8	and	and	CCONJ
fcis-9479	26	9	information	information	NOUN
fcis-9479	26	10	presented	present	VERB
fcis-9479	26	11	on	on	ADP
fcis-9479	26	12	social	social	ADJ
fcis-9479	26	13	media	medium	NOUN
fcis-9479	26	14	are	be	AUX
fcis-9479	26	15	characterized	characterize	VERB
fcis-9479	26	16	by	by	ADP
fcis-9479	26	17	diverse	diverse	ADJ
fcis-9479	26	18	modes	mode	NOUN
fcis-9479	26	19	and	and	CCONJ
fcis-9479	26	20	semantic	semantic	ADJ
fcis-9479	26	21	diversity[2	diversity[2	PROPN
fcis-9479	26	22	]	]	PUNCT
fcis-9479	26	23	.	.	PUNCT
fcis-9479	27	1	along	along	ADP
fcis-9479	27	2	with	with	ADP
fcis-9479	27	3	the	the	DET
fcis-9479	27	4	popularity	popularity	NOUN
fcis-9479	27	5	of	of	ADP
fcis-9479	27	6	mobile	mobile	ADJ
fcis-9479	27	7	devices	device	NOUN
fcis-9479	27	8	,	,	PUNCT
fcis-9479	27	9	news	news	NOUN
fcis-9479	27	10	is	be	AUX
fcis-9479	27	11	often	often	ADV
fcis-9479	27	12	published	publish	VERB
fcis-9479	27	13	and	and	CCONJ
fcis-9479	27	14	disseminated	disseminate	VERB
fcis-9479	27	15	rapidly	rapidly	ADV
fcis-9479	27	16	on	on	ADP
fcis-9479	27	17	social	social	ADJ
fcis-9479	27	18	media	medium	NOUN
fcis-9479	27	19	in	in	ADP
fcis-9479	27	20	the	the	DET
fcis-9479	27	21	form	form	NOUN
fcis-9479	27	22	of	of	ADP
fcis-9479	27	23	text	text	NOUN
fcis-9479	27	24	,	,	PUNCT
fcis-9479	27	25	images	image	NOUN
fcis-9479	27	26	,	,	PUNCT
fcis-9479	27	27	videos	video	NOUN
fcis-9479	27	28	and	and	CCONJ
fcis-9479	27	29	other	other	ADJ
fcis-9479	27	30	multimedia	multimedia	NOUN
fcis-9479	27	31	data	datum	NOUN
fcis-9479	27	32	.	.	PUNCT
fcis-9479	28	1	while	while	SCONJ
fcis-9479	28	2	this	this	DET
fcis-9479	28	3	kind	kind	NOUN
fcis-9479	28	4	of	of	ADP
fcis-9479	28	5	news	news	NOUN
fcis-9479	28	6	is	be	AUX
fcis-9479	28	7	convenient	convenient	ADJ
fcis-9479	28	8	for	for	SCONJ
fcis-9479	28	9	readers	reader	NOUN
fcis-9479	28	10	to	to	PART
fcis-9479	28	11	obtain	obtain	VERB
fcis-9479	28	12	and	and	CCONJ
fcis-9479	28	13	understand	understand	VERB
fcis-9479	28	14	information	information	NOUN
fcis-9479	28	15	,	,	PUNCT
fcis-9479	28	16	it	it	PRON
fcis-9479	28	17	also	also	ADV
fcis-9479	28	18	causes	cause	VERB
fcis-9479	28	19	the	the	DET
fcis-9479	28	20	traditional	traditional	ADJ
fcis-9479	28	21	technique	technique	NOUN
fcis-9479	28	22	of	of	ADP
fcis-9479	28	23	relying	rely	VERB
fcis-9479	28	24	solely	solely	ADV
fcis-9479	28	25	on	on	ADP
fcis-9479	28	26	text	text	NOUN
fcis-9479	28	27	analysis	analysis	NOUN
fcis-9479	28	28	for	for	ADP
fcis-9479	28	29	false	false	ADJ
fcis-9479	28	30	news	news	NOUN
fcis-9479	28	31	detection	detection	NOUN
fcis-9479	28	32	to	to	PART
fcis-9479	28	33	be	be	AUX
fcis-9479	28	34	no	no	ADV
fcis-9479	28	35	longer	long	ADV
fcis-9479	28	36	applicable	applicable	ADJ
fcis-9479	28	37	.	.	PUNCT
fcis-9479	29	1	another	another	DET
fcis-9479	29	2	example	example	NOUN
fcis-9479	29	3	is	be	AUX
fcis-9479	29	4	that	that	SCONJ
fcis-9479	29	5	various	various	ADJ
fcis-9479	29	6	types	type	NOUN
fcis-9479	29	7	of	of	ADP
fcis-9479	29	8	fake	fake	ADJ
fcis-9479	29	9	news	news	NOUN
fcis-9479	29	10	about	about	ADP
fcis-9479	29	11	the	the	DET
fcis-9479	29	12	location	location	NOUN
fcis-9479	29	13	and	and	CCONJ
fcis-9479	29	14	scale	scale	NOUN
fcis-9479	29	15	of	of	ADP
fcis-9479	29	16	exercise	exercise	NOUN
fcis-9479	29	17	in	in	ADP
fcis-9479	29	18	several	several	ADJ
fcis-9479	29	19	chinese	chinese	ADJ
fcis-9479	29	20	military	military	ADJ
fcis-9479	29	21	exercises	exercise	NOUN
fcis-9479	29	22	often	often	ADV
fcis-9479	29	23	spread	spread	VERB
fcis-9479	29	24	false	false	ADJ
fcis-9479	29	25	information	information	NOUN
fcis-9479	29	26	by	by	ADP
fcis-9479	29	27	stealing	steal	VERB
fcis-9479	29	28	irrelevant	irrelevant	ADJ
fcis-9479	29	29	pictures	picture	NOUN
fcis-9479	29	30	and	and	CCONJ
fcis-9479	29	31	videos	video	NOUN
fcis-9479	29	32	.	.	PUNCT
fcis-9479	30	1	1.2	1.2	NUM
fcis-9479	30	2	.	.	PUNCT
fcis-9479	30	3	status	status	NOUN
fcis-9479	30	4	of	of	ADP
fcis-9479	30	5	domestic	domestic	ADJ
fcis-9479	30	6	and	and	CCONJ
fcis-9479	30	7	international	international	ADJ
fcis-9479	30	8	research	research	NOUN
fcis-9479	30	9	as	as	ADP
fcis-9479	30	10	deep	deep	ADJ
fcis-9479	30	11	learning	learning	NOUN
fcis-9479	30	12	continues	continue	VERB
fcis-9479	30	13	to	to	PART
fcis-9479	30	14	make	make	VERB
fcis-9479	30	15	progress	progress	NOUN
fcis-9479	30	16	,	,	PUNCT
fcis-9479	30	17	more	more	ADJ
fcis-9479	30	18	and	and	CCONJ
fcis-9479	30	19	more	more	ADJ
fcis-9479	30	20	researchers	researcher	NOUN
fcis-9479	30	21	are	be	AUX
fcis-9479	30	22	studying	study	VERB
fcis-9479	30	23	fake	fake	ADJ
fcis-9479	30	24	news	news	NOUN
fcis-9479	30	25	detection	detection	NOUN
fcis-9479	30	26	based	base	VERB
fcis-9479	30	27	on	on	ADP
fcis-9479	30	28	neural	neural	ADJ
fcis-9479	30	29	networks	network	NOUN
fcis-9479	30	30	.	.	PUNCT
fcis-9479	31	1	there	there	PRON
fcis-9479	31	2	are	be	VERB
fcis-9479	31	3	many	many	ADJ
fcis-9479	31	4	branches	branch	NOUN
fcis-9479	31	5	in	in	ADP
fcis-9479	31	6	the	the	DET
fcis-9479	31	7	field	field	NOUN
fcis-9479	31	8	of	of	ADP
fcis-9479	31	9	fake	fake	ADJ
fcis-9479	31	10	news	news	NOUN
fcis-9479	31	11	detection	detection	NOUN
fcis-9479	31	12	,	,	PUNCT
fcis-9479	31	13	such	such	ADJ
fcis-9479	31	14	as	as	ADP
fcis-9479	31	15	unimodal	unimodal	NOUN
fcis-9479	31	16	-	-	PUNCT
fcis-9479	31	17	based	base	VERB
fcis-9479	31	18	fake	fake	ADJ
fcis-9479	31	19	news	news	NOUN
fcis-9479	31	20	detection	detection	NOUN
fcis-9479	31	21	,	,	PUNCT
fcis-9479	31	22	multimodal	multimodal	NOUN
fcis-9479	31	23	-	-	PUNCT
fcis-9479	31	24	based	base	VERB
fcis-9479	31	25	fake	fake	ADJ
fcis-9479	31	26	news	news	NOUN
fcis-9479	31	27	detection	detection	NOUN
fcis-9479	31	28	,	,	PUNCT
fcis-9479	31	29	etc	etc	X
fcis-9479	31	30	.	.	X
fcis-9479	31	31	unimodal	unimodal	NOUN
fcis-9479	31	32	-	-	PUNCT
fcis-9479	31	33	based	base	VERB
fcis-9479	31	34	fake	fake	ADJ
fcis-9479	31	35	news	news	NOUN
fcis-9479	31	36	detection	detection	NOUN
fcis-9479	31	37	means	mean	VERB
fcis-9479	31	38	that	that	SCONJ
fcis-9479	31	39	the	the	DET
fcis-9479	31	40	text	text	NOUN
fcis-9479	31	41	contains	contain	VERB
fcis-9479	31	42	only	only	ADV
fcis-9479	31	43	one	one	NUM
fcis-9479	31	44	of	of	ADP
fcis-9479	31	45	text	text	NOUN
fcis-9479	31	46	,	,	PUNCT
fcis-9479	31	47	picture	picture	NOUN
fcis-9479	31	48	or	or	CCONJ
fcis-9479	31	49	video	video	NOUN
fcis-9479	31	50	,	,	PUNCT
fcis-9479	31	51	multimodal	multimodal	NOUN
fcis-9479	31	52	-	-	PUNCT
fcis-9479	31	53	based	base	VERB
fcis-9479	31	54	fake	fake	ADJ
fcis-9479	31	55	news	news	NOUN
fcis-9479	31	56	detection	detection	NOUN
fcis-9479	31	57	refers	refer	VERB
fcis-9479	31	58	to	to	ADP
fcis-9479	31	59	multiple	multiple	ADJ
fcis-9479	31	60	types	type	NOUN
fcis-9479	31	61	of	of	ADP
fcis-9479	31	62	fake	fake	ADJ
fcis-9479	31	63	news	news	NOUN
fcis-9479	31	64	that	that	PRON
fcis-9479	31	65	contain	contain	VERB
fcis-9479	31	66	text	text	NOUN
fcis-9479	31	67	,	,	PUNCT
fcis-9479	31	68	pictures	picture	NOUN
fcis-9479	31	69	,	,	PUNCT
fcis-9479	31	70	videos	video	NOUN
fcis-9479	31	71	,	,	PUNCT
fcis-9479	31	72	voices	voice	NOUN
fcis-9479	31	73	,	,	PUNCT
fcis-9479	31	74	etc	etc	X
fcis-9479	31	75	.	.	X
fcis-9479	32	1	(	(	PUNCT
fcis-9479	32	2	1	1	X
fcis-9479	32	3	)	)	PUNCT
fcis-9479	32	4	unimodal	unimodal	NOUN
fcis-9479	32	5	-	-	PUNCT
fcis-9479	32	6	based	base	VERB
fcis-9479	32	7	method	method	NOUN
fcis-9479	32	8	for	for	ADP
fcis-9479	32	9	false	false	ADJ
fcis-9479	32	10	news	news	NOUN
fcis-9479	32	11	detection	detection	NOUN
fcis-9479	32	12	the	the	DET
fcis-9479	32	13	fake	fake	ADJ
fcis-9479	32	14	news	news	NOUN
fcis-9479	32	15	detection	detection	NOUN
fcis-9479	32	16	method[3	method[3	PROPN
fcis-9479	32	17	]	]	PUNCT
fcis-9479	32	18	based	base	VERB
fcis-9479	32	19	on	on	ADP
fcis-9479	32	20	single	single	ADJ
fcis-9479	32	21	modality	modality	NOUN
fcis-9479	32	22	mainly	mainly	ADV
fcis-9479	32	23	judges	judge	VERB
fcis-9479	32	24	the	the	DET
fcis-9479	32	25	authenticity	authenticity	NOUN
fcis-9479	32	26	of	of	ADP
fcis-9479	32	27	news	news	NOUN
fcis-9479	32	28	by	by	ADP
fcis-9479	32	29	extracting	extract	VERB
fcis-9479	32	30	text	text	NOUN
fcis-9479	32	31	features	feature	NOUN
fcis-9479	32	32	from	from	ADP
fcis-9479	32	33	news	news	NOUN
fcis-9479	32	34	text	text	NOUN
fcis-9479	32	35	content	content	NOUN
fcis-9479	32	36	or	or	CCONJ
fcis-9479	32	37	visual	visual	ADJ
fcis-9479	32	38	features	feature	NOUN
fcis-9479	32	39	from	from	ADP
fcis-9479	32	40	image	image	NOUN
fcis-9479	32	41	information	information	NOUN
fcis-9479	32	42	.	.	PUNCT
fcis-9479	33	1	for	for	ADP
fcis-9479	33	2	example	example	NOUN
fcis-9479	33	3	,	,	PUNCT
fcis-9479	33	4	castillo	castillo	PROPN
fcis-9479	33	5	et	et	PROPN
fcis-9479	33	6	al	al	PROPN
fcis-9479	33	7	.	.	PUNCT
fcis-9479	34	1	[	[	X
fcis-9479	34	2	3]used	3]used	NUM
fcis-9479	34	3	decision	decision	NOUN
fcis-9479	34	4	trees	tree	NOUN
fcis-9479	34	5	to	to	PART
fcis-9479	34	6	learn	learn	VERB
fcis-9479	34	7	thematic	thematic	ADJ
fcis-9479	34	8	features	feature	NOUN
fcis-9479	34	9	of	of	ADP
fcis-9479	34	10	news	news	NOUN
fcis-9479	34	11	text	text	NOUN
fcis-9479	34	12	content	content	NOUN
fcis-9479	34	13	for	for	ADP
fcis-9479	34	14	classification	classification	NOUN
fcis-9479	34	15	.	.	PUNCT
fcis-9479	35	1	the	the	DET
fcis-9479	35	2	model	model	NOUN
fcis-9479	35	3	proposed	propose	VERB
fcis-9479	35	4	by	by	ADP
fcis-9479	35	5	yu	yu	PROPN
fcis-9479	35	6	et	et	PROPN
fcis-9479	35	7	al	al	PROPN
fcis-9479	35	8	.	.	PUNCT
fcis-9479	36	1	[	[	X
fcis-9479	36	2	4]obtained	4]obtaine	VERB
fcis-9479	36	3	high	high	ADJ
fcis-9479	36	4	-	-	PUNCT
fcis-9479	36	5	level	level	NOUN
fcis-9479	36	6	interaction	interaction	NOUN
fcis-9479	36	7	features	feature	NOUN
fcis-9479	36	8	and	and	CCONJ
fcis-9479	36	9	key	key	ADJ
fcis-9479	36	10	features	feature	NOUN
fcis-9479	36	11	of	of	ADP
fcis-9479	36	12	relevant	relevant	ADJ
fcis-9479	36	13	posts	post	NOUN
fcis-9479	36	14	through	through	ADP
fcis-9479	36	15	convolutional	convolutional	ADJ
fcis-9479	36	16	neural	neural	ADJ
fcis-9479	36	17	networks.ma	networks.ma	NUM
fcis-9479	36	18	et	et	NOUN
fcis-9479	36	19	al.[5	al.[5	PROPN
fcis-9479	36	20	]	]	PUNCT
fcis-9479	36	21	used	use	VERB
fcis-9479	36	22	recurrent	recurrent	ADJ
fcis-9479	36	23	neural	neural	ADJ
fcis-9479	36	24	networks	network	NOUN
fcis-9479	36	25	to	to	PART
fcis-9479	36	26	learn	learn	VERB
fcis-9479	36	27	latent	latent	NOUN
fcis-9479	36	28	features	feature	NOUN
fcis-9479	36	29	of	of	ADP
fcis-9479	36	30	news	news	NOUN
fcis-9479	36	31	text	text	NOUN
fcis-9479	36	32	content	content	NOUN
fcis-9479	36	33	.	.	PUNCT
fcis-9479	37	1	in	in	ADP
fcis-9479	37	2	mvnn[6	mvnn[6	PROPN
fcis-9479	37	3	]	]	PUNCT
fcis-9479	37	4	,	,	PUNCT
fcis-9479	37	5	the	the	DET
fcis-9479	37	6	authors	author	NOUN
fcis-9479	37	7	used	use	VERB
fcis-9479	37	8	a	a	DET
fcis-9479	37	9	multiregion	multiregion	NOUN
fcis-9479	37	10	visual	visual	ADJ
fcis-9479	37	11	neural	neural	ADJ
fcis-9479	37	12	network	network	NOUN
fcis-9479	37	13	for	for	ADP
fcis-9479	37	14	fake	fake	ADJ
fcis-9479	37	15	news	news	NOUN
fcis-9479	37	16	detection	detection	NOUN
fcis-9479	37	17	by	by	ADP
fcis-9479	37	18	targeting	target	VERB
fcis-9479	37	19	the	the	DET
fcis-9479	37	20	rich	rich	ADJ
fcis-9479	37	21	visual	visual	ADJ
fcis-9479	37	22	information	information	NOUN
fcis-9479	37	23	in	in	ADP
fcis-9479	37	24	different	different	ADJ
fcis-9479	37	25	pixel	pixel	NOUN
fcis-9479	37	26	regions	region	NOUN
fcis-9479	37	27	.	.	PUNCT
fcis-9479	38	1	(	(	PUNCT
fcis-9479	38	2	2	2	X
fcis-9479	38	3	)	)	PUNCT
fcis-9479	38	4	multimodal	multimodal	NOUN
fcis-9479	38	5	-	-	PUNCT
fcis-9479	38	6	based	base	VERB
fcis-9479	38	7	method	method	NOUN
fcis-9479	38	8	for	for	ADP
fcis-9479	38	9	fake	fake	ADJ
fcis-9479	38	10	news	news	NOUN
fcis-9479	38	11	detection	detection	NOUN
fcis-9479	38	12	106	106	NUM
fcis-9479	38	13	fake	fake	ADJ
fcis-9479	38	14	news	news	NOUN
fcis-9479	38	15	detection	detection	NOUN
fcis-9479	38	16	based	base	VERB
fcis-9479	38	17	on	on	ADP
fcis-9479	38	18	multimodality	multimodality	NOUN
fcis-9479	38	19	has	have	AUX
fcis-9479	38	20	attracted	attract	VERB
fcis-9479	38	21	a	a	DET
fcis-9479	38	22	lot	lot	NOUN
fcis-9479	38	23	of	of	ADP
fcis-9479	38	24	attention	attention	NOUN
fcis-9479	38	25	in	in	ADP
fcis-9479	38	26	recent	recent	ADJ
fcis-9479	38	27	years	year	NOUN
fcis-9479	38	28	.	.	PUNCT
fcis-9479	39	1	some	some	PRON
fcis-9479	39	2	of	of	ADP
fcis-9479	39	3	these	these	DET
fcis-9479	39	4	methods	method	NOUN
fcis-9479	39	5	concatenate	concatenate	VERB
fcis-9479	39	6	the	the	DET
fcis-9479	39	7	textual	textual	ADJ
fcis-9479	39	8	features	feature	NOUN
fcis-9479	39	9	in	in	ADP
fcis-9479	39	10	the	the	DET
fcis-9479	39	11	post	post	NOUN
fcis-9479	39	12	with	with	ADP
fcis-9479	39	13	the	the	DET
fcis-9479	39	14	visual	visual	ADJ
fcis-9479	39	15	features	feature	NOUN
fcis-9479	39	16	of	of	ADP
fcis-9479	39	17	the	the	DET
fcis-9479	39	18	images	image	NOUN
fcis-9479	39	19	in	in	ADP
fcis-9479	39	20	the	the	DET
fcis-9479	39	21	post[7	post[7	NOUN
fcis-9479	39	22	]	]	PUNCT
fcis-9479	39	23	.	.	PUNCT
fcis-9479	40	1	however	however	ADV
fcis-9479	40	2	,	,	PUNCT
fcis-9479	40	3	this	this	DET
fcis-9479	40	4	approach	approach	NOUN
fcis-9479	40	5	requires	require	VERB
fcis-9479	40	6	manual	manual	ADJ
fcis-9479	40	7	feature	feature	NOUN
fcis-9479	40	8	engineering	engineering	NOUN
fcis-9479	40	9	on	on	ADP
fcis-9479	40	10	the	the	DET
fcis-9479	40	11	one	one	NUM
fcis-9479	40	12	hand	hand	NOUN
fcis-9479	40	13	,	,	PUNCT
fcis-9479	40	14	and	and	CCONJ
fcis-9479	40	15	is	be	AUX
fcis-9479	40	16	not	not	PART
fcis-9479	40	17	able	able	ADJ
fcis-9479	40	18	to	to	PART
fcis-9479	40	19	effectively	effectively	ADV
fcis-9479	40	20	obtain	obtain	VERB
fcis-9479	40	21	complex	complex	ADJ
fcis-9479	40	22	semantic	semantic	ADJ
fcis-9479	40	23	representations	representation	NOUN
fcis-9479	40	24	in	in	ADP
fcis-9479	40	25	images	image	NOUN
fcis-9479	40	26	on	on	ADP
fcis-9479	40	27	the	the	DET
fcis-9479	40	28	other	other	ADJ
fcis-9479	40	29	hand	hand	NOUN
fcis-9479	40	30	.	.	PUNCT
fcis-9479	41	1	at	at	ADP
fcis-9479	41	2	present	present	ADJ
fcis-9479	41	3	,	,	PUNCT
fcis-9479	41	4	due	due	ADP
fcis-9479	41	5	to	to	ADP
fcis-9479	41	6	the	the	DET
fcis-9479	41	7	excellent	excellent	ADJ
fcis-9479	41	8	performance	performance	NOUN
fcis-9479	41	9	of	of	ADP
fcis-9479	41	10	deep	deep	ADJ
fcis-9479	41	11	neural	neural	ADJ
fcis-9479	41	12	network	network	NOUN
fcis-9479	41	13	(	(	PUNCT
fcis-9479	41	14	dnn	dnn	PROPN
fcis-9479	41	15	)	)	PUNCT
fcis-9479	41	16	in	in	ADP
fcis-9479	41	17	nonlinear	nonlinear	ADJ
fcis-9479	41	18	representation	representation	NOUN
fcis-9479	41	19	learning[8	learning[8	PROPN
fcis-9479	41	20	]	]	PUNCT
fcis-9479	41	21	,	,	PUNCT
fcis-9479	41	22	many	many	ADJ
fcis-9479	41	23	multimodal	multimodal	NOUN
fcis-9479	41	24	representation	representation	NOUN
fcis-9479	41	25	learning	learning	NOUN
fcis-9479	41	26	methods	method	NOUN
fcis-9479	41	27	use	use	VERB
fcis-9479	41	28	deep	deep	ADJ
fcis-9479	41	29	learning	learning	NOUN
fcis-9479	41	30	mechanism	mechanism	NOUN
fcis-9479	41	31	to	to	PART
fcis-9479	41	32	learn	learn	VERB
fcis-9479	41	33	feature	feature	NOUN
fcis-9479	41	34	representation	representation	NOUN
fcis-9479	41	35	,	,	PUNCT
fcis-9479	41	36	thereby	thereby	ADV
fcis-9479	41	37	improving	improve	VERB
fcis-9479	41	38	the	the	DET
fcis-9479	41	39	ability	ability	NOUN
fcis-9479	41	40	of	of	ADP
fcis-9479	41	41	fake	fake	ADJ
fcis-9479	41	42	news	news	NOUN
fcis-9479	41	43	detection	detection	NOUN
fcis-9479	41	44	.	.	PUNCT
fcis-9479	42	1	jin	jin	NOUN
fcis-9479	42	2	et	et	PROPN
fcis-9479	42	3	al	al	PROPN
fcis-9479	42	4	.	.	PUNCT
fcis-9479	43	1	[	[	X
fcis-9479	43	2	9]proposed	9]proposed	X
fcis-9479	43	3	a	a	DET
fcis-9479	43	4	method	method	NOUN
fcis-9479	43	5	based	base	VERB
fcis-9479	43	6	on	on	ADP
fcis-9479	43	7	deep	deep	ADJ
fcis-9479	43	8	learning	learning	NOUN
fcis-9479	43	9	,	,	PUNCT
fcis-9479	43	10	which	which	PRON
fcis-9479	43	11	can	can	AUX
fcis-9479	43	12	learn	learn	VERB
fcis-9479	43	13	the	the	DET
fcis-9479	43	14	multimodal	multimodal	NOUN
fcis-9479	43	15	content	content	NOUN
fcis-9479	43	16	and	and	CCONJ
fcis-9479	43	17	social	social	ADJ
fcis-9479	43	18	information	information	NOUN
fcis-9479	43	19	of	of	ADP
fcis-9479	43	20	news	news	NOUN
fcis-9479	43	21	posts	post	NOUN
fcis-9479	43	22	,	,	PUNCT
fcis-9479	43	23	and	and	CCONJ
fcis-9479	43	24	then	then	ADV
fcis-9479	43	25	used	use	VERB
fcis-9479	43	26	the	the	DET
fcis-9479	43	27	attention	attention	NOUN
fcis-9479	43	28	mechanism	mechanism	NOUN
fcis-9479	43	29	to	to	PART
fcis-9479	43	30	fuse	fuse	VERB
fcis-9479	43	31	the	the	DET
fcis-9479	43	32	multimodal	multimodal	NOUN
fcis-9479	43	33	features	feature	NOUN
fcis-9479	43	34	.	.	PUNCT
fcis-9479	44	1	in	in	ADP
fcis-9479	44	2	eann[10	eann[10	NOUN
fcis-9479	44	3	]	]	PUNCT
fcis-9479	44	4	,	,	PUNCT
fcis-9479	44	5	the	the	DET
fcis-9479	44	6	authors	author	NOUN
fcis-9479	44	7	learned	learn	VERB
fcis-9479	44	8	the	the	DET
fcis-9479	44	9	invariant	invariant	ADJ
fcis-9479	44	10	features	feature	NOUN
fcis-9479	44	11	of	of	ADP
fcis-9479	44	12	events	event	NOUN
fcis-9479	44	13	through	through	ADP
fcis-9479	44	14	an	an	DET
fcis-9479	44	15	adversarial	adversarial	ADJ
fcis-9479	44	16	network	network	NOUN
fcis-9479	44	17	containing	contain	VERB
fcis-9479	44	18	a	a	DET
fcis-9479	44	19	multimodal	multimodal	ADJ
fcis-9479	44	20	feature	feature	NOUN
fcis-9479	44	21	extractor	extractor	NOUN
fcis-9479	44	22	to	to	PART
fcis-9479	44	23	obtain	obtain	VERB
fcis-9479	44	24	multimodal	multimodal	ADJ
fcis-9479	44	25	features	feature	NOUN
fcis-9479	44	26	of	of	ADP
fcis-9479	44	27	each	each	DET
fcis-9479	44	28	news	news	NOUN
fcis-9479	44	29	item	item	NOUN
fcis-9479	44	30	for	for	ADP
fcis-9479	44	31	fake	fake	ADJ
fcis-9479	44	32	news	news	NOUN
fcis-9479	44	33	detection	detection	NOUN
fcis-9479	44	34	.	.	PUNCT
fcis-9479	45	1	in	in	ADP
fcis-9479	45	2	mvae[11	mvae[11	PROPN
fcis-9479	45	3	]	]	PUNCT
fcis-9479	45	4	,	,	PUNCT
fcis-9479	45	5	khattar	khattar	NOUN
fcis-9479	45	6	et	et	PROPN
fcis-9479	45	7	al	al	PROPN
fcis-9479	45	8	.	.	PROPN
fcis-9479	45	9	used	use	VERB
fcis-9479	45	10	a	a	DET
fcis-9479	45	11	multimodal	multimodal	ADJ
fcis-9479	45	12	variational	variational	ADJ
fcis-9479	45	13	self	self	NOUN
fcis-9479	45	14	-	-	PUNCT
fcis-9479	45	15	encoder	encoder	NOUN
fcis-9479	45	16	for	for	ADP
fcis-9479	45	17	fake	fake	ADJ
fcis-9479	45	18	news	news	NOUN
fcis-9479	45	19	identification	identification	NOUN
fcis-9479	45	20	,	,	PUNCT
fcis-9479	45	21	where	where	SCONJ
fcis-9479	45	22	the	the	DET
fcis-9479	45	23	multimodal	multimodal	NOUN
fcis-9479	45	24	features	feature	NOUN
fcis-9479	45	25	of	of	ADP
fcis-9479	45	26	a	a	DET
fcis-9479	45	27	post	post	NOUN
fcis-9479	45	28	were	be	AUX
fcis-9479	45	29	fed	feed	VERB
fcis-9479	45	30	into	into	ADP
fcis-9479	45	31	a	a	DET
fcis-9479	45	32	bimodal	bimodal	NOUN
fcis-9479	45	33	variational	variational	ADJ
fcis-9479	45	34	self	self	NOUN
fcis-9479	45	35	-	-	PUNCT
fcis-9479	45	36	encoder	encoder	NOUN
fcis-9479	45	37	to	to	PART
fcis-9479	45	38	obtain	obtain	VERB
fcis-9479	45	39	a	a	DET
fcis-9479	45	40	multimodal	multimodal	ADJ
fcis-9479	45	41	feature	feature	NOUN
fcis-9479	45	42	representation	representation	NOUN
fcis-9479	45	43	of	of	ADP
fcis-9479	45	44	the	the	DET
fcis-9479	45	45	news	news	NOUN
fcis-9479	45	46	.	.	PUNCT
fcis-9479	46	1	cui	cui	NOUN
fcis-9479	46	2	et	et	PROPN
fcis-9479	46	3	al	al	PROPN
fcis-9479	46	4	.	.	PUNCT
fcis-9479	47	1	[	[	X
fcis-9479	47	2	12]proposed	12]proposed	NUM
fcis-9479	47	3	an	an	DET
fcis-9479	47	4	end	end	NOUN
fcis-9479	47	5	-	-	PUNCT
fcis-9479	47	6	to	to	ADP
fcis-9479	47	7	-	-	PUNCT
fcis-9479	47	8	end	end	NOUN
fcis-9479	47	9	deep	deep	ADJ
fcis-9479	47	10	embedding	embed	VERB
fcis-9479	47	11	framework	framework	NOUN
fcis-9479	47	12	for	for	ADP
fcis-9479	47	13	fake	fake	ADJ
fcis-9479	47	14	news	news	NOUN
fcis-9479	47	15	detection	detection	NOUN
fcis-9479	47	16	,	,	PUNCT
fcis-9479	47	17	in	in	ADP
fcis-9479	47	18	which	which	PRON
fcis-9479	47	19	the	the	DET
fcis-9479	47	20	latent	latent	ADJ
fcis-9479	47	21	sentiment	sentiment	NOUN
fcis-9479	47	22	of	of	ADP
fcis-9479	47	23	the	the	DET
fcis-9479	47	24	post	post	NOUN
fcis-9479	47	25	publisher	publisher	NOUN
fcis-9479	47	26	was	be	AUX
fcis-9479	47	27	used	use	VERB
fcis-9479	47	28	to	to	PART
fcis-9479	47	29	distinguish	distinguish	VERB
fcis-9479	47	30	fake	fake	ADJ
fcis-9479	47	31	news	news	NOUN
fcis-9479	47	32	.	.	PUNCT
fcis-9479	48	1	spotfake[13	spotfake[13	PROPN
fcis-9479	48	2	]	]	PUNCT
fcis-9479	48	3	used	use	VERB
fcis-9479	48	4	a	a	DET
fcis-9479	48	5	pre	pre	ADJ
fcis-9479	48	6	-	-	ADJ
fcis-9479	48	7	trained	train	VERB
fcis-9479	48	8	bert	bert	NOUN
fcis-9479	49	1	[	[	X
fcis-9479	49	2	14]model	14]model	NUM
fcis-9479	49	3	to	to	PART
fcis-9479	49	4	learn	learn	VERB
fcis-9479	49	5	text	text	NOUN
fcis-9479	49	6	features	feature	NOUN
fcis-9479	49	7	of	of	ADP
fcis-9479	49	8	news	news	NOUN
fcis-9479	49	9	posts	post	NOUN
fcis-9479	49	10	and	and	CCONJ
fcis-9479	49	11	a	a	DET
fcis-9479	49	12	vgg-19	vgg-19	PROPN
fcis-9479	49	13	model	model	NOUN
fcis-9479	49	14	pre	pre	VERB
fcis-9479	49	15	-	-	VERB
fcis-9479	49	16	trained	train	VERB
fcis-9479	49	17	on	on	ADP
fcis-9479	49	18	imagenet	imagenet	NOUN
fcis-9479	49	19	[	[	X
fcis-9479	49	20	15	15	NUM
fcis-9479	49	21	]	]	PUNCT
fcis-9479	49	22	to	to	PART
fcis-9479	49	23	extract	extract	VERB
fcis-9479	49	24	image	image	NOUN
fcis-9479	49	25	features	feature	NOUN
fcis-9479	49	26	.	.	PUNCT
fcis-9479	50	1	spotfake+	spotfake+	PUNCT
fcis-9479	51	1	[	[	X
fcis-9479	51	2	16]was	16]was	NUM
fcis-9479	51	3	an	an	DET
fcis-9479	51	4	improved	improved	ADJ
fcis-9479	51	5	version	version	NOUN
fcis-9479	51	6	of	of	ADP
fcis-9479	51	7	spotfake	spotfake	NOUN
fcis-9479	51	8	that	that	PRON
fcis-9479	51	9	used	use	VERB
fcis-9479	51	10	a	a	DET
fcis-9479	51	11	modified	modify	VERB
fcis-9479	51	12	version	version	NOUN
fcis-9479	51	13	of	of	ADP
fcis-9479	51	14	bert	bert	PROPN
fcis-9479	51	15	,	,	PUNCT
fcis-9479	51	16	the	the	DET
fcis-9479	51	17	xlnet	xlnet	PROPN
fcis-9479	52	1	[	[	X
fcis-9479	52	2	17]model	17]model	NUM
fcis-9479	52	3	,	,	PUNCT
fcis-9479	52	4	to	to	PART
fcis-9479	52	5	extract	extract	VERB
fcis-9479	52	6	text	text	NOUN
fcis-9479	52	7	features	feature	NOUN
fcis-9479	52	8	based	base	VERB
fcis-9479	52	9	on	on	ADP
fcis-9479	52	10	spotfake	spotfake	NOUN
fcis-9479	52	11	.	.	PUNCT
fcis-9479	53	1	while	while	SCONJ
fcis-9479	53	2	learning	learn	VERB
fcis-9479	53	3	news	news	NOUN
fcis-9479	53	4	text	text	NOUN
fcis-9479	53	5	features	feature	NOUN
fcis-9479	53	6	and	and	CCONJ
fcis-9479	53	7	visual	visual	ADJ
fcis-9479	53	8	information	information	NOUN
fcis-9479	53	9	,	,	PUNCT
fcis-9479	53	10	the	the	DET
fcis-9479	53	11	safe[18	safe[18	PROPN
fcis-9479	53	12	]	]	PUNCT
fcis-9479	53	13	model	model	NOUN
fcis-9479	53	14	also	also	ADV
fcis-9479	53	15	learned	learn	VERB
fcis-9479	53	16	the	the	DET
fcis-9479	53	17	intrinsic	intrinsic	ADJ
fcis-9479	53	18	connection	connection	NOUN
fcis-9479	53	19	between	between	ADP
fcis-9479	53	20	text	text	NOUN
fcis-9479	53	21	content	content	NOUN
fcis-9479	53	22	and	and	CCONJ
fcis-9479	53	23	vision	vision	NOUN
fcis-9479	53	24	to	to	PART
fcis-9479	53	25	predict	predict	VERB
fcis-9479	53	26	fake	fake	ADJ
fcis-9479	53	27	news	news	NOUN
fcis-9479	53	28	.	.	PUNCT
fcis-9479	54	1	in	in	ADP
fcis-9479	54	2	m	m	NOUN
fcis-9479	54	3	-	-	NOUN
fcis-9479	54	4	gcn[19	gcn[19	NOUN
fcis-9479	54	5	]	]	PUNCT
fcis-9479	54	6	,	,	PUNCT
fcis-9479	54	7	the	the	DET
fcis-9479	54	8	author	author	NOUN
fcis-9479	54	9	focused	focus	VERB
fcis-9479	54	10	on	on	ADP
fcis-9479	54	11	distinguishing	distinguish	VERB
fcis-9479	54	12	different	different	ADJ
fcis-9479	54	13	degrees	degree	NOUN
fcis-9479	54	14	of	of	ADP
fcis-9479	54	15	fake	fake	ADJ
fcis-9479	54	16	news	news	NOUN
fcis-9479	54	17	according	accord	VERB
fcis-9479	54	18	to	to	ADP
fcis-9479	54	19	the	the	DET
fcis-9479	54	20	similarity	similarity	NOUN
fcis-9479	54	21	between	between	ADP
fcis-9479	54	22	news	news	NOUN
fcis-9479	54	23	,	,	PUNCT
fcis-9479	54	24	which	which	PRON
fcis-9479	54	25	used	use	VERB
fcis-9479	54	26	gcn	gcn	NOUN
fcis-9479	54	27	modules	module	NOUN
fcis-9479	54	28	of	of	ADP
fcis-9479	54	29	different	different	ADJ
fcis-9479	54	30	depths	depth	NOUN
fcis-9479	54	31	to	to	PART
fcis-9479	54	32	extract	extract	VERB
fcis-9479	54	33	domain	domain	NOUN
fcis-9479	54	34	information	information	NOUN
fcis-9479	54	35	of	of	ADP
fcis-9479	54	36	different	different	ADJ
fcis-9479	54	37	scales	scale	NOUN
fcis-9479	54	38	,	,	PUNCT
fcis-9479	54	39	and	and	CCONJ
fcis-9479	54	40	fused	fuse	VERB
fcis-9479	54	41	these	these	DET
fcis-9479	54	42	features	feature	NOUN
fcis-9479	54	43	through	through	ADP
fcis-9479	54	44	the	the	DET
fcis-9479	54	45	attention	attention	NOUN
fcis-9479	54	46	mechanism	mechanism	NOUN
fcis-9479	54	47	.	.	PUNCT
fcis-9479	55	1	2	2	X
fcis-9479	55	2	.	.	X
fcis-9479	55	3	research	research	NOUN
fcis-9479	55	4	methods	method	NOUN
fcis-9479	55	5	2.1	2.1	NUM
fcis-9479	55	6	.	.	PUNCT
fcis-9479	56	1	cnn	cnn	PROPN
fcis-9479	56	2	and	and	CCONJ
fcis-9479	56	3	lstm	lstm	PROPN
fcis-9479	56	4	model	model	NOUN
fcis-9479	56	5	fusion	fusion	NOUN
fcis-9479	56	6	before	before	ADP
fcis-9479	56	7	training	train	VERB
fcis-9479	56	8	the	the	DET
fcis-9479	56	9	model	model	NOUN
fcis-9479	56	10	,	,	PUNCT
fcis-9479	56	11	the	the	DET
fcis-9479	56	12	text	text	NOUN
fcis-9479	56	13	was	be	AUX
fcis-9479	56	14	processed	process	VERB
fcis-9479	56	15	,	,	PUNCT
fcis-9479	56	16	such	such	ADJ
fcis-9479	56	17	as	as	ADP
fcis-9479	56	18	removing	remove	VERB
fcis-9479	56	19	spaces	space	NOUN
fcis-9479	56	20	and	and	CCONJ
fcis-9479	56	21	keeping	keep	VERB
fcis-9479	56	22	only	only	ADV
fcis-9479	56	23	the	the	DET
fcis-9479	56	24	text	text	NOUN
fcis-9479	56	25	,	,	PUNCT
fcis-9479	56	26	then	then	ADV
fcis-9479	56	27	the	the	DET
fcis-9479	56	28	text	text	NOUN
fcis-9479	56	29	was	be	AUX
fcis-9479	56	30	divided	divide	VERB
fcis-9479	56	31	into	into	ADP
fcis-9479	56	32	words	word	NOUN
fcis-9479	56	33	and	and	CCONJ
fcis-9479	56	34	deactivated	deactivate	VERB
fcis-9479	56	35	words	word	NOUN
fcis-9479	56	36	,	,	PUNCT
fcis-9479	56	37	after	after	ADP
fcis-9479	56	38	the	the	DET
fcis-9479	56	39	division	division	NOUN
fcis-9479	56	40	of	of	ADP
fcis-9479	56	41	words	word	NOUN
fcis-9479	56	42	,	,	PUNCT
fcis-9479	56	43	the	the	DET
fcis-9479	56	44	training	training	NOUN
fcis-9479	56	45	vocabulary	vocabulary	NOUN
fcis-9479	56	46	was	be	AUX
fcis-9479	56	47	tracked	track	VERB
fcis-9479	56	48	,	,	PUNCT
fcis-9479	56	49	all	all	DET
fcis-9479	56	50	the	the	DET
fcis-9479	56	51	text	text	NOUN
fcis-9479	56	52	was	be	AUX
fcis-9479	56	53	traversed	traverse	VERB
fcis-9479	56	54	and	and	CCONJ
fcis-9479	56	55	the	the	DET
fcis-9479	56	56	words	word	NOUN
fcis-9479	56	57	were	be	AUX
fcis-9479	56	58	counted	count	VERB
fcis-9479	56	59	,	,	PUNCT
fcis-9479	56	60	then	then	ADV
fcis-9479	56	61	the	the	DET
fcis-9479	56	62	text	text	NOUN
fcis-9479	56	63	was	be	AUX
fcis-9479	56	64	converted	convert	VERB
fcis-9479	56	65	into	into	ADP
fcis-9479	56	66	a	a	DET
fcis-9479	56	67	vector	vector	NOUN
fcis-9479	56	68	sequence	sequence	NOUN
fcis-9479	56	69	,	,	PUNCT
fcis-9479	56	70	and	and	CCONJ
fcis-9479	56	71	filled	fill	VERB
fcis-9479	56	72	the	the	DET
fcis-9479	56	73	vector	vector	NOUN
fcis-9479	56	74	sequence	sequence	NOUN
fcis-9479	56	75	according	accord	VERB
fcis-9479	56	76	to	to	ADP
fcis-9479	56	77	the	the	DET
fcis-9479	56	78	maximum	maximum	ADJ
fcis-9479	56	79	length	length	NOUN
fcis-9479	56	80	of	of	ADP
fcis-9479	56	81	the	the	DET
fcis-9479	56	82	text	text	NOUN
fcis-9479	56	83	,	,	PUNCT
fcis-9479	56	84	making	make	VERB
fcis-9479	56	85	the	the	DET
fcis-9479	56	86	sequence	sequence	NOUN
fcis-9479	56	87	of	of	ADP
fcis-9479	56	88	vectors	vector	NOUN
fcis-9479	56	89	to	to	ADP
fcis-9479	56	90	uniform	uniform	ADJ
fcis-9479	56	91	length	length	NOUN
fcis-9479	56	92	.	.	PUNCT
fcis-9479	57	1	then	then	ADV
fcis-9479	57	2	word	word	NOUN
fcis-9479	57	3	vector	vector	NOUN
fcis-9479	57	4	training	training	NOUN
fcis-9479	57	5	was	be	AUX
fcis-9479	57	6	performed	perform	VERB
fcis-9479	57	7	using	use	VERB
fcis-9479	57	8	word2vec_model	word2vec_model	NOUN
fcis-9479	57	9	.	.	PUNCT
fcis-9479	58	1	the	the	DET
fcis-9479	58	2	attention	attention	NOUN
fcis-9479	58	3	mechanism	mechanism	NOUN
fcis-9479	58	4	was	be	AUX
fcis-9479	58	5	introduced	introduce	VERB
fcis-9479	58	6	in	in	ADP
fcis-9479	58	7	the	the	DET
fcis-9479	58	8	model	model	NOUN
fcis-9479	58	9	to	to	PART
fcis-9479	58	10	capture	capture	VERB
fcis-9479	58	11	the	the	DET
fcis-9479	58	12	key	key	ADJ
fcis-9479	58	13	points	point	NOUN
fcis-9479	58	14	from	from	ADP
fcis-9479	58	15	the	the	DET
fcis-9479	58	16	longer	long	ADJ
fcis-9479	58	17	text	text	NOUN
fcis-9479	58	18	without	without	ADP
fcis-9479	58	19	losing	lose	VERB
fcis-9479	58	20	important	important	ADJ
fcis-9479	58	21	information	information	NOUN
fcis-9479	58	22	.	.	PUNCT
fcis-9479	59	1	after	after	ADP
fcis-9479	59	2	processing	process	VERB
fcis-9479	59	3	the	the	DET
fcis-9479	59	4	text	text	NOUN
fcis-9479	59	5	data	datum	NOUN
fcis-9479	59	6	,	,	PUNCT
fcis-9479	59	7	the	the	DET
fcis-9479	59	8	training	training	NOUN
fcis-9479	59	9	was	be	AUX
fcis-9479	59	10	performed	perform	VERB
fcis-9479	59	11	with	with	ADP
fcis-9479	59	12	cnn	cnn	PROPN
fcis-9479	59	13	and	and	CCONJ
fcis-9479	59	14	lstm	lstm	NOUN
fcis-9479	59	15	respectively	respectively	ADV
fcis-9479	59	16	,	,	PUNCT
fcis-9479	59	17	finally	finally	ADV
fcis-9479	59	18	,	,	PUNCT
fcis-9479	59	19	lr	lr	PROPN
fcis-9479	59	20	was	be	AUX
fcis-9479	59	21	used	use	VERB
fcis-9479	59	22	to	to	PART
fcis-9479	59	23	try	try	VERB
fcis-9479	59	24	to	to	PART
fcis-9479	59	25	get	get	VERB
fcis-9479	59	26	the	the	DET
fcis-9479	59	27	fusion	fusion	NOUN
fcis-9479	59	28	coefficient	coefficient	NOUN
fcis-9479	59	29	of	of	ADP
fcis-9479	59	30	the	the	DET
fcis-9479	59	31	two	two	NUM
fcis-9479	59	32	models	model	NOUN
fcis-9479	59	33	of	of	ADP
fcis-9479	59	34	cnn	cnn	PROPN
fcis-9479	59	35	and	and	CCONJ
fcis-9479	59	36	lstm	lstm	PROPN
fcis-9479	59	37	,	,	PUNCT
fcis-9479	59	38	here	here	ADV
fcis-9479	59	39	was	be	AUX
fcis-9479	59	40	the	the	DET
fcis-9479	59	41	regression	regression	NOUN
fcis-9479	59	42	coefficient	coefficient	NOUN
fcis-9479	59	43	of	of	ADP
fcis-9479	59	44	lr	lr	NOUN
fcis-9479	59	45	,	,	PUNCT
fcis-9479	59	46	and	and	CCONJ
fcis-9479	59	47	then	then	ADV
fcis-9479	59	48	the	the	DET
fcis-9479	59	49	best	good	ADJ
fcis-9479	59	50	threshold	threshold	NOUN
fcis-9479	59	51	was	be	AUX
fcis-9479	59	52	calculated	calculate	VERB
fcis-9479	59	53	on	on	ADP
fcis-9479	59	54	the	the	DET
fcis-9479	59	55	validation	validation	NOUN
fcis-9479	59	56	set	set	VERB
fcis-9479	59	57	with	with	ADP
fcis-9479	59	58	the	the	DET
fcis-9479	59	59	fused	fuse	VERB
fcis-9479	59	60	model	model	NOUN
fcis-9479	59	61	.	.	PUNCT
fcis-9479	60	1	2.2	2.2	NUM
fcis-9479	60	2	.	.	PUNCT
fcis-9479	60	3	lightgbm	lightgbm	VERB
fcis-9479	60	4	and	and	CCONJ
fcis-9479	60	5	xgboost	xgboost	NOUN
fcis-9479	60	6	models	model	NOUN
fcis-9479	60	7	segment	segment	VERB
fcis-9479	60	8	the	the	DET
fcis-9479	60	9	text	text	NOUN
fcis-9479	60	10	data	datum	NOUN
fcis-9479	60	11	first	first	ADV
fcis-9479	60	12	,	,	PUNCT
fcis-9479	60	13	fill	fill	VERB
fcis-9479	60	14	in	in	ADP
fcis-9479	60	15	the	the	DET
fcis-9479	60	16	empty	empty	ADJ
fcis-9479	60	17	values	value	NOUN
fcis-9479	60	18	after	after	ADP
fcis-9479	60	19	segmentation	segmentation	NOUN
fcis-9479	60	20	,	,	PUNCT
fcis-9479	60	21	and	and	CCONJ
fcis-9479	60	22	use	use	VERB
fcis-9479	60	23	the	the	DET
fcis-9479	60	24	tf	tf	PROPN
fcis-9479	60	25	-	-	PUNCT
fcis-9479	60	26	idf	idf	PROPN
fcis-9479	60	27	algorithm	algorithm	NOUN
fcis-9479	60	28	to	to	PART
fcis-9479	60	29	extract	extract	VERB
fcis-9479	60	30	the	the	DET
fcis-9479	60	31	feature	feature	NOUN
fcis-9479	60	32	vector	vector	NOUN
fcis-9479	60	33	of	of	ADP
fcis-9479	60	34	the	the	DET
fcis-9479	60	35	news	news	NOUN
fcis-9479	60	36	text	text	NOUN
fcis-9479	60	37	.	.	PUNCT
fcis-9479	61	1	the	the	DET
fcis-9479	61	2	tf	tf	PROPN
fcis-9479	61	3	-	-	PUNCT
fcis-9479	61	4	idf	idf	PROPN
fcis-9479	61	5	algorithm	algorithm	NOUN
fcis-9479	61	6	is	be	AUX
fcis-9479	61	7	one	one	NUM
fcis-9479	61	8	of	of	ADP
fcis-9479	61	9	the	the	DET
fcis-9479	61	10	important	important	ADJ
fcis-9479	61	11	algorithms	algorithm	NOUN
fcis-9479	61	12	for	for	ADP
fcis-9479	61	13	extracting	extract	VERB
fcis-9479	61	14	feature	feature	NOUN
fcis-9479	61	15	word	word	NOUN
fcis-9479	61	16	vectors	vector	NOUN
fcis-9479	61	17	,	,	PUNCT
fcis-9479	61	18	and	and	CCONJ
fcis-9479	61	19	it	it	PRON
fcis-9479	61	20	is	be	AUX
fcis-9479	61	21	also	also	ADV
fcis-9479	61	22	one	one	NUM
fcis-9479	61	23	of	of	ADP
fcis-9479	61	24	the	the	DET
fcis-9479	61	25	main	main	ADJ
fcis-9479	61	26	technologies	technology	NOUN
fcis-9479	61	27	for	for	ADP
fcis-9479	61	28	generating	generate	VERB
fcis-9479	61	29	word	word	NOUN
fcis-9479	61	30	vectors	vector	NOUN
fcis-9479	61	31	.	.	PUNCT
fcis-9479	62	1	the	the	DET
fcis-9479	62	2	tf	tf	PROPN
fcis-9479	62	3	-	-	PUNCT
fcis-9479	62	4	idf	idf	PROPN
fcis-9479	62	5	algorithm	algorithm	NOUN
fcis-9479	62	6	statistically	statistically	ADV
fcis-9479	62	7	evaluates	evaluate	VERB
fcis-9479	62	8	the	the	DET
fcis-9479	62	9	importance	importance	NOUN
fcis-9479	62	10	of	of	ADP
fcis-9479	62	11	a	a	DET
fcis-9479	62	12	word	word	NOUN
fcis-9479	62	13	to	to	ADP
fcis-9479	62	14	a	a	DET
fcis-9479	62	15	document	document	NOUN
fcis-9479	62	16	or	or	CCONJ
fcis-9479	62	17	other	other	ADJ
fcis-9479	62	18	documents	document	NOUN
fcis-9479	62	19	in	in	ADP
fcis-9479	62	20	the	the	DET
fcis-9479	62	21	corpus	corpus	NOUN
fcis-9479	62	22	to	to	PART
fcis-9479	62	23	determine	determine	VERB
fcis-9479	62	24	the	the	DET
fcis-9479	62	25	feature	feature	NOUN
fcis-9479	62	26	words	word	NOUN
fcis-9479	62	27	of	of	ADP
fcis-9479	62	28	the	the	DET
fcis-9479	62	29	document	document	NOUN
fcis-9479	62	30	.	.	PUNCT
fcis-9479	63	1	the	the	DET
fcis-9479	63	2	basic	basic	ADJ
fcis-9479	63	3	idea	idea	NOUN
fcis-9479	63	4	:	:	PUNCT
fcis-9479	63	5	if	if	SCONJ
fcis-9479	63	6	a	a	DET
fcis-9479	63	7	word	word	NOUN
fcis-9479	63	8	appears	appear	VERB
fcis-9479	63	9	frequently	frequently	ADV
fcis-9479	63	10	in	in	ADP
fcis-9479	63	11	a	a	DET
fcis-9479	63	12	document	document	NOUN
fcis-9479	63	13	,	,	PUNCT
fcis-9479	63	14	but	but	CCONJ
fcis-9479	63	15	appears	appear	VERB
fcis-9479	63	16	infrequently	infrequently	ADV
fcis-9479	63	17	in	in	ADP
fcis-9479	63	18	other	other	ADJ
fcis-9479	63	19	documents	document	NOUN
fcis-9479	63	20	in	in	ADP
fcis-9479	63	21	the	the	DET
fcis-9479	63	22	corpus	corpus	NOUN
fcis-9479	63	23	,	,	PUNCT
fcis-9479	63	24	it	it	PRON
fcis-9479	63	25	is	be	AUX
fcis-9479	63	26	determined	determine	VERB
fcis-9479	63	27	that	that	SCONJ
fcis-9479	63	28	the	the	DET
fcis-9479	63	29	word	word	NOUN
fcis-9479	63	30	can	can	AUX
fcis-9479	63	31	be	be	AUX
fcis-9479	63	32	used	use	VERB
fcis-9479	63	33	as	as	ADP
fcis-9479	63	34	a	a	DET
fcis-9479	63	35	feature	feature	NOUN
fcis-9479	63	36	word	word	NOUN
fcis-9479	63	37	of	of	ADP
fcis-9479	63	38	the	the	DET
fcis-9479	63	39	document	document	NOUN
fcis-9479	63	40	to	to	ADP
fcis-9479	63	41	some	some	DET
fcis-9479	63	42	extent	extent	NOUN
fcis-9479	63	43	.	.	PUNCT
fcis-9479	64	1	it	it	PRON
fcis-9479	64	2	has	have	VERB
fcis-9479	64	3	the	the	DET
fcis-9479	64	4	ability	ability	NOUN
fcis-9479	64	5	to	to	PART
fcis-9479	64	6	distinguish	distinguish	VERB
fcis-9479	64	7	categories	category	NOUN
fcis-9479	64	8	and	and	CCONJ
fcis-9479	64	9	can	can	AUX
fcis-9479	64	10	be	be	AUX
fcis-9479	64	11	used	use	VERB
fcis-9479	64	12	as	as	ADP
fcis-9479	64	13	the	the	DET
fcis-9479	64	14	basis	basis	NOUN
fcis-9479	64	15	for	for	ADP
fcis-9479	64	16	classification	classification	NOUN
fcis-9479	64	17	.	.	PUNCT
fcis-9479	65	1	the	the	DET
fcis-9479	65	2	tf	tf	PROPN
fcis-9479	65	3	-	-	PUNCT
fcis-9479	65	4	idf	idf	PROPN
fcis-9479	65	5	algorithm	algorithm	NOUN
fcis-9479	65	6	is	be	AUX
fcis-9479	65	7	divided	divide	VERB
fcis-9479	65	8	into	into	ADP
fcis-9479	65	9	tf	tf	PROPN
fcis-9479	65	10	(	(	PUNCT
fcis-9479	65	11	term	term	NOUN
fcis-9479	65	12	frequency	frequency	NOUN
fcis-9479	65	13	)	)	PUNCT
fcis-9479	65	14	algorithm	algorithm	NOUN
fcis-9479	65	15	and	and	CCONJ
fcis-9479	65	16	idf	idf	PROPN
fcis-9479	65	17	(	(	PUNCT
fcis-9479	65	18	inverse	inverse	NOUN
fcis-9479	65	19	document	document	NOUN
fcis-9479	65	20	frequency	frequency	NOUN
fcis-9479	65	21	)	)	PUNCT
fcis-9479	65	22	algorithm	algorithm	NOUN
fcis-9479	65	23	,	,	PUNCT
fcis-9479	65	24	where	where	SCONJ
fcis-9479	65	25	the	the	DET
fcis-9479	65	26	tf	tf	PROPN
fcis-9479	65	27	algorithm	algorithm	NOUN
fcis-9479	65	28	represents	represent	VERB
fcis-9479	65	29	the	the	DET
fcis-9479	65	30	ratio	ratio	NOUN
fcis-9479	65	31	of	of	ADP
fcis-9479	65	32	the	the	DET
fcis-9479	65	33	count	count	NOUN
fcis-9479	65	34	of	of	ADP
fcis-9479	65	35	a	a	DET
fcis-9479	65	36	specific	specific	ADJ
fcis-9479	65	37	word	word	NOUN
fcis-9479	65	38	to	to	ADP
fcis-9479	65	39	the	the	DET
fcis-9479	65	40	total	total	ADJ
fcis-9479	65	41	number	number	NOUN
fcis-9479	65	42	of	of	ADP
fcis-9479	65	43	words	word	NOUN
fcis-9479	65	44	in	in	ADP
fcis-9479	65	45	the	the	DET
fcis-9479	65	46	document	document	NOUN
fcis-9479	65	47	,	,	PUNCT
fcis-9479	65	48	representing	represent	VERB
fcis-9479	65	49	the	the	DET
fcis-9479	65	50	frequency	frequency	NOUN
fcis-9479	65	51	of	of	ADP
fcis-9479	65	52	a	a	DET
fcis-9479	65	53	specific	specific	ADJ
fcis-9479	65	54	word	word	NOUN
fcis-9479	65	55	in	in	ADP
fcis-9479	65	56	the	the	DET
fcis-9479	65	57	document	document	NOUN
fcis-9479	65	58	.	.	PUNCT
fcis-9479	66	1	the	the	DET
fcis-9479	66	2	idf	idf	PROPN
fcis-9479	66	3	algorithm	algorithm	NOUN
fcis-9479	66	4	represents	represent	VERB
fcis-9479	66	5	the	the	DET
fcis-9479	66	6	logarithm	logarithm	NOUN
fcis-9479	66	7	of	of	ADP
fcis-9479	66	8	the	the	DET
fcis-9479	66	9	ratio	ratio	NOUN
fcis-9479	66	10	of	of	ADP
fcis-9479	66	11	the	the	DET
fcis-9479	66	12	total	total	ADJ
fcis-9479	66	13	number	number	NOUN
fcis-9479	66	14	of	of	ADP
fcis-9479	66	15	documents	document	NOUN
fcis-9479	66	16	in	in	ADP
fcis-9479	66	17	the	the	DET
fcis-9479	66	18	corpus	corpus	NOUN
fcis-9479	66	19	to	to	ADP
fcis-9479	66	20	the	the	DET
fcis-9479	66	21	number	number	NOUN
fcis-9479	66	22	of	of	ADP
fcis-9479	66	23	documents	document	NOUN
fcis-9479	66	24	in	in	ADP
fcis-9479	66	25	which	which	PRON
fcis-9479	66	26	a	a	DET
fcis-9479	66	27	particular	particular	ADJ
fcis-9479	66	28	word	word	NOUN
fcis-9479	66	29	occurs	occur	VERB
fcis-9479	66	30	in	in	ADP
fcis-9479	66	31	the	the	DET
fcis-9479	66	32	corpus	corpus	NOUN
fcis-9479	66	33	.	.	PUNCT
fcis-9479	67	1	then	then	ADV
fcis-9479	67	2	set	set	VERB
fcis-9479	67	3	the	the	DET
fcis-9479	67	4	model	model	NOUN
fcis-9479	67	5	parameters	parameter	NOUN
fcis-9479	67	6	respectively	respectively	ADV
fcis-9479	67	7	,	,	PUNCT
fcis-9479	67	8	and	and	CCONJ
fcis-9479	67	9	finally	finally	ADV
fcis-9479	67	10	train	train	VERB
fcis-9479	67	11	the	the	DET
fcis-9479	67	12	model	model	NOUN
fcis-9479	67	13	.	.	PUNCT
fcis-9479	68	1	3	3	X
fcis-9479	68	2	.	.	X
fcis-9479	68	3	introduction	introduction	NOUN
fcis-9479	68	4	of	of	ADP
fcis-9479	68	5	the	the	DET
fcis-9479	68	6	experimental	experimental	ADJ
fcis-9479	68	7	data	datum	NOUN
fcis-9479	68	8	set	set	VERB
fcis-9479	68	9	in	in	ADP
fcis-9479	68	10	the	the	DET
fcis-9479	68	11	false	false	ADJ
fcis-9479	68	12	news	news	NOUN
fcis-9479	68	13	text	text	NOUN
fcis-9479	68	14	detection	detection	NOUN
fcis-9479	68	15	task	task	NOUN
fcis-9479	68	16	,	,	PUNCT
fcis-9479	68	17	the	the	DET
fcis-9479	68	18	training	training	NOUN
fcis-9479	68	19	set	set	NOUN
fcis-9479	68	20	contains	contain	VERB
fcis-9479	68	21	a	a	DET
fcis-9479	68	22	total	total	NOUN
fcis-9479	68	23	of	of	ADP
fcis-9479	68	24	38,471	38,471	NUM
fcis-9479	68	25	news	news	NOUN
fcis-9479	68	26	items	item	NOUN
fcis-9479	68	27	,	,	PUNCT
fcis-9479	68	28	including	include	VERB
fcis-9479	68	29	19,186	19,186	NUM
fcis-9479	68	30	real	real	ADJ
fcis-9479	68	31	news	news	NOUN
fcis-9479	68	32	items	item	NOUN
fcis-9479	68	33	and	and	CCONJ
fcis-9479	68	34	19,285	19,285	NUM
fcis-9479	68	35	false	false	ADJ
fcis-9479	68	36	news	news	NOUN
fcis-9479	68	37	items	item	NOUN
fcis-9479	68	38	.	.	PUNCT
fcis-9479	69	1	each	each	DET
fcis-9479	69	2	data	datum	NOUN
fcis-9479	69	3	consists	consist	VERB
fcis-9479	69	4	of	of	ADP
fcis-9479	69	5	three	three	NUM
fcis-9479	69	6	elements	element	NOUN
fcis-9479	69	7	[	[	X
fcis-9479	69	8	i	i	X
fcis-9479	69	9	d	d	PROPN
fcis-9479	69	10	,	,	PUNCT
fcis-9479	69	11	text	text	NOUN
fcis-9479	69	12	,	,	PUNCT
fcis-9479	69	13	label	label	NOUN
fcis-9479	69	14	]	]	PUNCT
fcis-9479	69	15	,	,	PUNCT
fcis-9479	69	16	where	where	SCONJ
fcis-9479	69	17	i	i	PRON
fcis-9479	69	18	d	d	PROPN
fcis-9479	69	19	is	be	AUX
fcis-9479	69	20	the	the	DET
fcis-9479	69	21	unique	unique	ADJ
fcis-9479	69	22	i	i	PRON
fcis-9479	69	23	d	d	PROPN
fcis-9479	69	24	of	of	ADP
fcis-9479	69	25	each	each	DET
fcis-9479	69	26	data	datum	NOUN
fcis-9479	69	27	,	,	PUNCT
fcis-9479	69	28	which	which	PRON
fcis-9479	69	29	uniquely	uniquely	ADV
fcis-9479	69	30	characterizes	characterize	VERB
fcis-9479	69	31	a	a	DET
fcis-9479	69	32	news	news	NOUN
fcis-9479	69	33	,	,	PUNCT
fcis-9479	69	34	text	text	NOUN
fcis-9479	69	35	is	be	AUX
fcis-9479	69	36	the	the	DET
fcis-9479	69	37	chinese	chinese	ADJ
fcis-9479	69	38	news	news	NOUN
fcis-9479	69	39	text	text	NOUN
fcis-9479	69	40	,	,	PUNCT
fcis-9479	69	41	label	label	NOUN
fcis-9479	69	42	is	be	AUX
fcis-9479	69	43	represented	represent	VERB
fcis-9479	69	44	by	by	ADP
fcis-9479	69	45	0	0	NUM
fcis-9479	69	46	and	and	CCONJ
fcis-9479	69	47	1	1	NUM
fcis-9479	69	48	,	,	PUNCT
fcis-9479	69	49	0	0	NUM
fcis-9479	69	50	means	mean	VERB
fcis-9479	69	51	real	real	ADJ
fcis-9479	69	52	news	news	NOUN
fcis-9479	69	53	,	,	PUNCT
fcis-9479	69	54	1	1	NUM
fcis-9479	69	55	means	mean	VERB
fcis-9479	69	56	false	false	ADJ
fcis-9479	69	57	news	news	NOUN
fcis-9479	69	58	.	.	PUNCT
fcis-9479	70	1	the	the	DET
fcis-9479	70	2	testing	testing	NOUN
fcis-9479	70	3	dataset	dataset	VERB
fcis-9479	70	4	needs	need	VERB
fcis-9479	70	5	to	to	PART
fcis-9479	70	6	be	be	AUX
fcis-9479	70	7	submitted	submit	VERB
fcis-9479	70	8	to	to	ADP
fcis-9479	70	9	the	the	DET
fcis-9479	70	10	backend	backend	NOUN
fcis-9479	70	11	for	for	ADP
fcis-9479	70	12	determination	determination	NOUN
fcis-9479	70	13	.	.	PUNCT
fcis-9479	71	1	analysis	analysis	NOUN
fcis-9479	71	2	of	of	ADP
fcis-9479	71	3	the	the	DET
fcis-9479	71	4	dataset	dataset	NOUN
fcis-9479	71	5	yields	yield	NOUN
fcis-9479	71	6	the	the	DET
fcis-9479	71	7	results	result	NOUN
fcis-9479	71	8	shown	show	VERB
fcis-9479	71	9	in	in	ADP
fcis-9479	71	10	table	table	NOUN
fcis-9479	71	11	1	1	NUM
fcis-9479	71	12	.	.	PUNCT
fcis-9479	71	13	table	table	NOUN
fcis-9479	71	14	1	1	NUM
fcis-9479	71	15	.	.	PUNCT
fcis-9479	71	16	dataset	dataset	VERB
fcis-9479	71	17	analysis	analysis	NOUN
fcis-9479	71	18	maximum	maximum	ADJ
fcis-9479	71	19	sentence	sentence	NOUN
fcis-9479	71	20	length	length	NOUN
fcis-9479	71	21	minimum	minimum	NOUN
fcis-9479	71	22	sentence	sentence	NOUN
fcis-9479	71	23	length	length	NOUN
fcis-9479	71	24	average	average	ADJ
fcis-9479	71	25	data	datum	NOUN
fcis-9479	71	26	length	length	NOUN
fcis-9479	71	27	number	number	NOUN
fcis-9479	71	28	of	of	ADP
fcis-9479	71	29	words	word	NOUN
fcis-9479	71	30	after	after	ADP
fcis-9479	71	31	the	the	DET
fcis-9479	71	32	participle	participle	NOUN
fcis-9479	71	33	929	929	NUM
fcis-9479	71	34	0	0	NUM
fcis-9479	71	35	50.883678615060695	50.883678615060695	NUM
fcis-9479	71	36	73915	73915	NUM
fcis-9479	71	37	4	4	NUM
fcis-9479	71	38	.	.	PUNCT
fcis-9479	71	39	experiment	experiment	NOUN
fcis-9479	71	40	and	and	CCONJ
fcis-9479	71	41	results	result	VERB
fcis-9479	71	42	4.1	4.1	NUM
fcis-9479	71	43	.	.	PUNCT
fcis-9479	72	1	experimental	experimental	ADJ
fcis-9479	72	2	environment	environment	NOUN
fcis-9479	72	3	(	(	PUNCT
fcis-9479	72	4	1	1	X
fcis-9479	72	5	)	)	PUNCT
fcis-9479	72	6	experimental	experimental	ADJ
fcis-9479	72	7	hardware	hardware	NOUN
fcis-9479	72	8	environment	environment	NOUN
fcis-9479	72	9	the	the	DET
fcis-9479	72	10	laptop	laptop	NOUN
fcis-9479	72	11	processor	processor	NOUN
fcis-9479	72	12	used	use	VERB
fcis-9479	72	13	in	in	ADP
fcis-9479	72	14	the	the	DET
fcis-9479	72	15	experiment	experiment	NOUN
fcis-9479	72	16	is	be	AUX
fcis-9479	72	17	i7	i7	NOUN
fcis-9479	72	18	-	-	PUNCT
fcis-9479	72	19	8550u	8550u	NUM
fcis-9479	72	20	,	,	PUNCT
fcis-9479	72	21	and	and	CCONJ
fcis-9479	72	22	the	the	DET
fcis-9479	72	23	memory	memory	NOUN
fcis-9479	72	24	is	be	AUX
fcis-9479	72	25	8.00	8.00	NUM
fcis-9479	72	26	gb	gb	NOUN
fcis-9479	72	27	.	.	PUNCT
fcis-9479	73	1	the	the	DET
fcis-9479	73	2	experimental	experimental	ADJ
fcis-9479	73	3	hardware	hardware	NOUN
fcis-9479	73	4	environment	environment	NOUN
fcis-9479	73	5	is	be	AUX
fcis-9479	73	6	shown	show	VERB
fcis-9479	73	7	in	in	ADP
fcis-9479	73	8	table	table	NOUN
fcis-9479	73	9	2	2	NUM
fcis-9479	73	10	.	.	PUNCT
fcis-9479	73	11	table	table	NOUN
fcis-9479	73	12	2	2	NUM
fcis-9479	73	13	.	.	PUNCT
fcis-9479	73	14	experimental	experimental	ADJ
fcis-9479	73	15	environment	environment	NOUN
fcis-9479	73	16	equipment	equipment	NOUN
fcis-9479	73	17	name	name	NOUN
fcis-9479	73	18	desktop-9c2ckr8	desktop-9c2ckr8	NOUN
fcis-9479	73	19	processor	processor	NOUN
fcis-9479	73	20	intel(r	intel(r	PROPN
fcis-9479	73	21	)	)	PUNCT
fcis-9479	73	22	core	core	NOUN
fcis-9479	73	23	(	(	PUNCT
fcis-9479	73	24	tm	tm	NOUN
fcis-9479	73	25	)	)	PUNCT
fcis-9479	73	26	i7	i7	ADJ
fcis-9479	73	27	-	-	PUNCT
fcis-9479	73	28	8550u	8550u	NOUN
fcis-9479	73	29	cpu	cpu	NOUN
fcis-9479	73	30	ram	ram	VERB
fcis-9479	73	31	8.00	8.00	NUM
fcis-9479	73	32	gb	gb	NOUN
fcis-9479	73	33	(	(	PUNCT
fcis-9479	73	34	7.9	7.9	NUM
fcis-9479	73	35	gb	gb	ADV
fcis-9479	73	36	available	available	ADJ
fcis-9479	73	37	)	)	PUNCT
fcis-9479	73	38	(	(	PUNCT
fcis-9479	73	39	2	2	X
fcis-9479	73	40	)	)	PUNCT
fcis-9479	73	41	experimental	experimental	ADJ
fcis-9479	73	42	software	software	NOUN
fcis-9479	73	43	environment	environment	NOUN
fcis-9479	73	44	the	the	DET
fcis-9479	73	45	experimental	experimental	ADJ
fcis-9479	73	46	code	code	NOUN
fcis-9479	73	47	in	in	ADP
fcis-9479	73	48	this	this	DET
fcis-9479	73	49	paper	paper	NOUN
fcis-9479	73	50	is	be	AUX
fcis-9479	73	51	completed	complete	VERB
fcis-9479	73	52	in	in	ADP
fcis-9479	73	53	python	python	NOUN
fcis-9479	73	54	language	language	NOUN
fcis-9479	73	55	,	,	PUNCT
fcis-9479	73	56	and	and	CCONJ
fcis-9479	73	57	the	the	DET
fcis-9479	73	58	python	python	PROPN
fcis-9479	73	59	version	version	NOUN
fcis-9479	73	60	used	use	VERB
fcis-9479	73	61	is	be	AUX
fcis-9479	73	62	3.7.4	3.7.4	NUM
fcis-9479	73	63	.	.	PUNCT
fcis-9479	74	1	during	during	ADP
fcis-9479	74	2	the	the	DET
fcis-9479	74	3	experiment	experiment	NOUN
fcis-9479	74	4	,	,	PUNCT
fcis-9479	74	5	pandas	panda	NOUN
fcis-9479	74	6	,	,	PUNCT
fcis-9479	74	7	matplotlib	matplotlib	PROPN
fcis-9479	74	8	,	,	PUNCT
fcis-9479	74	9	numpy	numpy	NOUN
fcis-9479	74	10	,	,	PUNCT
fcis-9479	74	11	tqdm	tqdm	ADJ
fcis-9479	74	12	,	,	PUNCT
fcis-9479	74	13	jieba	jieba	PROPN
fcis-9479	74	14	,	,	PUNCT
fcis-9479	74	15	scikitlearn	scikitlearn	PROPN
fcis-9479	74	16	,	,	PUNCT
fcis-9479	74	17	keras	keras	PROPN
fcis-9479	74	18	,	,	PUNCT
fcis-9479	74	19	tensorflow	tensorflow	NOUN
fcis-9479	74	20	,	,	PUNCT
fcis-9479	74	21	xgboost	xgboost	ADV
fcis-9479	74	22	,	,	PUNCT
fcis-9479	74	23	scipy	scipy	NOUN
fcis-9479	74	24	,	,	PUNCT
fcis-9479	74	25	lightgbm	lightgbm	ADJ
fcis-9479	74	26	packages	package	NOUN
fcis-9479	74	27	were	be	AUX
fcis-9479	74	28	called	call	VERB
fcis-9479	74	29	,	,	PUNCT
fcis-9479	74	30	and	and	CCONJ
fcis-9479	74	31	their	their	PRON
fcis-9479	74	32	version	version	NOUN
fcis-9479	74	33	numbers	number	NOUN
fcis-9479	74	34	are	be	AUX
fcis-9479	74	35	pandas	panda	NOUN
fcis-9479	74	36	(	(	PUNCT
fcis-9479	74	37	1.3.5	1.3.5	NUM
fcis-9479	74	38	)	)	PUNCT
fcis-9479	74	39	,	,	PUNCT
fcis-9479	74	40	matplotlib	matplotlib	PROPN
fcis-9479	74	41	(	(	PUNCT
fcis-9479	74	42	3.1.1	3.1.1	NUM
fcis-9479	74	43	)	)	PUNCT
fcis-9479	74	44	,	,	PUNCT
fcis-9479	74	45	numpy	numpy	NOUN
fcis-9479	74	46	(	(	PUNCT
fcis-9479	74	47	1.21.5	1.21.5	NUM
fcis-9479	74	48	)	)	PUNCT
fcis-9479	74	49	,	,	PUNCT
fcis-9479	74	50	tqdm	tqdm	NOUN
fcis-9479	74	51	(	(	PUNCT
fcis-9479	74	52	4.62.3	4.62.3	NOUN
fcis-9479	74	53	)	)	PUNCT
fcis-9479	74	54	,	,	PUNCT
fcis-9479	74	55	jieba	jieba	X
fcis-9479	74	56	(	(	PUNCT
fcis-9479	74	57	0.42.1	0.42.1	NUM
fcis-9479	74	58	)	)	PUNCT
fcis-9479	74	59	,	,	PUNCT
fcis-9479	74	60	scikit	scikit	NOUN
fcis-9479	74	61	-	-	PUNCT
fcis-9479	74	62	learn	learn	VERB
fcis-9479	74	63	(	(	PUNCT
fcis-9479	74	64	0.21.3	0.21.3	NUM
fcis-9479	74	65	)	)	PUNCT
fcis-9479	74	66	,	,	PUNCT
fcis-9479	74	67	keras	keras	PROPN
fcis-9479	74	68	(	(	PUNCT
fcis-9479	74	69	2.2.5	2.2.5	NUM
fcis-9479	74	70	)	)	PUNCT
fcis-9479	74	71	,	,	PUNCT
fcis-9479	74	72	tensorflow	tensorflow	NOUN
fcis-9479	74	73	(	(	PUNCT
fcis-9479	74	74	1.15.0	1.15.0	NUM
fcis-9479	74	75	)	)	PUNCT
fcis-9479	74	76	,	,	PUNCT
fcis-9479	74	77	xgboost	xgboost	X
fcis-9479	74	78	(	(	PUNCT
fcis-9479	74	79	1.6.1	1.6.1	NUM
fcis-9479	74	80	)	)	PUNCT
fcis-9479	74	81	,	,	PUNCT
fcis-9479	74	82	scipy	scipy	PROPN
fcis-9479	74	83	(	(	PUNCT
fcis-9479	74	84	1.3.1	1.3.1	NUM
fcis-9479	74	85	)	)	PUNCT
fcis-9479	74	86	,	,	PUNCT
fcis-9479	74	87	lightgbm	lightgbm	X
fcis-9479	74	88	(	(	PUNCT
fcis-9479	74	89	3.3.2	3.3.2	NUM
fcis-9479	74	90	)	)	PUNCT
fcis-9479	74	91	.	.	PUNCT
fcis-9479	75	1	107	107	NUM
fcis-9479	75	2	table	table	NOUN
fcis-9479	75	3	3	3	NUM
fcis-9479	75	4	.	.	PUNCT
fcis-9479	75	5	library	library	NOUN
fcis-9479	75	6	name	name	NOUN
fcis-9479	75	7	and	and	CCONJ
fcis-9479	75	8	version	version	NOUN
fcis-9479	75	9	library	library	NOUN
fcis-9479	75	10	name	name	NOUN
fcis-9479	75	11	version	version	PROPN
fcis-9479	75	12	python	python	NOUN
fcis-9479	75	13	3.7.4	3.7.4	NUM
fcis-9479	75	14	pandas	panda	NOUN
fcis-9479	75	15	1.3.5	1.3.5	NUM
fcis-9479	75	16	matplotlib	matplotlib	PROPN
fcis-9479	75	17	3.1.1	3.1.1	NUM
fcis-9479	75	18	numpy	numpy	NOUN
fcis-9479	75	19	1.21.5	1.21.5	NUM
fcis-9479	75	20	tqdm	tqdm	NOUN
fcis-9479	75	21	4.62.3	4.62.3	PROPN
fcis-9479	75	22	jieba	jieba	NOUN
fcis-9479	75	23	0.42.1	0.42.1	NUM
fcis-9479	75	24	scikit	scikit	NOUN
fcis-9479	75	25	-	-	PUNCT
fcis-9479	75	26	learn	learn	VERB
fcis-9479	75	27	0.21.3	0.21.3	NUM
fcis-9479	75	28	keras	keras	PROPN
fcis-9479	75	29	2.2.5	2.2.5	NUM
fcis-9479	75	30	tensorflow	tensorflow	NOUN
fcis-9479	75	31	1.15.0	1.15.0	NUM
fcis-9479	76	1	xgboost	xgboost	ADP
fcis-9479	76	2	1.6.1	1.6.1	NUM
fcis-9479	76	3	scipy	scipy	NOUN
fcis-9479	76	4	1.3.1	1.3.1	NUM
fcis-9479	76	5	lightgbm	lightgbm	VERB
fcis-9479	76	6	3.3.2	3.3.2	NUM
fcis-9479	76	7	4.2	4.2	NUM
fcis-9479	76	8	.	.	PUNCT
fcis-9479	77	1	experimental	experimental	ADJ
fcis-9479	77	2	procedure	procedure	NOUN
fcis-9479	77	3	4.2.1	4.2.1	NUM
fcis-9479	77	4	.	.	PUNCT
fcis-9479	77	5	model	model	NOUN
fcis-9479	77	6	training	train	VERB
fcis-9479	77	7	the	the	DET
fcis-9479	77	8	training	training	NOUN
fcis-9479	77	9	process	process	NOUN
fcis-9479	77	10	of	of	ADP
fcis-9479	77	11	cnn	cnn	PROPN
fcis-9479	77	12	and	and	CCONJ
fcis-9479	77	13	lstm	lstm	PROPN
fcis-9479	77	14	model	model	NOUN
fcis-9479	77	15	fusion	fusion	NOUN
fcis-9479	77	16	is	be	AUX
fcis-9479	77	17	shown	show	VERB
fcis-9479	77	18	in	in	ADP
fcis-9479	77	19	figure	figure	NOUN
fcis-9479	77	20	1	1	NUM
fcis-9479	77	21	and	and	CCONJ
fcis-9479	77	22	figure	figure	VERB
fcis-9479	77	23	2	2	NUM
fcis-9479	77	24	.	.	PUNCT
fcis-9479	78	1	it	it	PRON
fcis-9479	78	2	can	can	AUX
fcis-9479	78	3	be	be	AUX
fcis-9479	78	4	seen	see	VERB
fcis-9479	78	5	from	from	ADP
fcis-9479	78	6	figure	figure	NOUN
fcis-9479	78	7	1	1	NUM
fcis-9479	78	8	that	that	SCONJ
fcis-9479	78	9	it	it	PRON
fcis-9479	78	10	took	take	VERB
fcis-9479	78	11	2.0396049999999377	2.0396049999999377	NUM
fcis-9479	78	12	seconds	second	NOUN
fcis-9479	78	13	to	to	PART
fcis-9479	78	14	import	import	VERB
fcis-9479	78	15	word2vec	word2vec	X
fcis-9479	78	16	during	during	ADP
fcis-9479	78	17	the	the	DET
fcis-9479	78	18	training	training	NOUN
fcis-9479	78	19	of	of	ADP
fcis-9479	78	20	the	the	DET
fcis-9479	78	21	word	word	NOUN
fcis-9479	78	22	vector	vector	NOUN
fcis-9479	78	23	model	model	NOUN
fcis-9479	78	24	.	.	PUNCT
fcis-9479	79	1	the	the	DET
fcis-9479	79	2	model	model	NOUN
fcis-9479	79	3	was	be	AUX
fcis-9479	79	4	trained	train	VERB
fcis-9479	79	5	in	in	ADP
fcis-9479	79	6	two	two	NUM
fcis-9479	79	7	batches	batch	NOUN
fcis-9479	79	8	.	.	PUNCT
fcis-9479	80	1	and	and	CCONJ
fcis-9479	80	2	it	it	PRON
fcis-9479	80	3	can	can	AUX
fcis-9479	80	4	be	be	AUX
fcis-9479	80	5	seen	see	VERB
fcis-9479	80	6	from	from	ADP
fcis-9479	80	7	figure	figure	NOUN
fcis-9479	80	8	2	2	NUM
fcis-9479	80	9	that	that	SCONJ
fcis-9479	80	10	the	the	DET
fcis-9479	80	11	f1	f1	PROPN
fcis-9479	80	12	scores	score	NOUN
fcis-9479	80	13	on	on	ADP
fcis-9479	80	14	the	the	DET
fcis-9479	80	15	verification	verification	NOUN
fcis-9479	80	16	set	set	NOUN
fcis-9479	80	17	were	be	AUX
fcis-9479	80	18	0.9429	0.9429	NUM
fcis-9479	80	19	and	and	CCONJ
fcis-9479	80	20	0.9637	0.9637	NUM
fcis-9479	80	21	respectively	respectively	ADV
fcis-9479	80	22	.	.	PUNCT
fcis-9479	81	1	figure	figure	NOUN
fcis-9479	81	2	1	1	NUM
fcis-9479	81	3	.	.	PUNCT
fcis-9479	81	4	model	model	NOUN
fcis-9479	81	5	training	training	NOUN
fcis-9479	81	6	process	process	NOUN
fcis-9479	81	7	figure	figure	NOUN
fcis-9479	81	8	2	2	NUM
fcis-9479	81	9	.	.	PUNCT
fcis-9479	82	1	the	the	DET
fcis-9479	82	2	f1	f1	PROPN
fcis-9479	82	3	values	value	NOUN
fcis-9479	82	4	in	in	ADP
fcis-9479	82	5	model	model	NOUN
fcis-9479	82	6	training	training	NOUN
fcis-9479	82	7	it	it	PRON
fcis-9479	82	8	can	can	AUX
fcis-9479	82	9	be	be	AUX
fcis-9479	82	10	seen	see	VERB
fcis-9479	82	11	from	from	ADP
fcis-9479	82	12	figure	figure	NOUN
fcis-9479	82	13	3	3	NUM
fcis-9479	82	14	that	that	SCONJ
fcis-9479	82	15	the	the	DET
fcis-9479	82	16	number	number	NOUN
fcis-9479	82	17	of	of	ADP
fcis-9479	82	18	positive	positive	ADJ
fcis-9479	82	19	samples	sample	NOUN
fcis-9479	82	20	used	use	VERB
fcis-9479	82	21	in	in	ADP
fcis-9479	82	22	the	the	DET
fcis-9479	82	23	experiment	experiment	NOUN
fcis-9479	82	24	is	be	AUX
fcis-9479	82	25	15673	15673	NUM
fcis-9479	82	26	,	,	PUNCT
fcis-9479	82	27	and	and	CCONJ
fcis-9479	82	28	the	the	DET
fcis-9479	82	29	number	number	NOUN
fcis-9479	82	30	of	of	ADP
fcis-9479	82	31	negative	negative	ADJ
fcis-9479	82	32	samples	sample	NOUN
fcis-9479	82	33	is	be	AUX
fcis-9479	82	34	15487	15487	NUM
fcis-9479	82	35	.	.	PUNCT
fcis-9479	83	1	during	during	ADP
fcis-9479	83	2	the	the	DET
fcis-9479	83	3	model	model	NOUN
fcis-9479	83	4	training	training	NOUN
fcis-9479	83	5	process	process	NOUN
fcis-9479	83	6	,	,	PUNCT
fcis-9479	83	7	31160	31160	NUM
fcis-9479	83	8	pieces	piece	NOUN
fcis-9479	83	9	of	of	ADP
fcis-9479	83	10	data	datum	NOUN
fcis-9479	83	11	were	be	AUX
fcis-9479	83	12	used	use	VERB
fcis-9479	83	13	for	for	ADP
fcis-9479	83	14	training	training	NOUN
fcis-9479	83	15	,	,	PUNCT
fcis-9479	83	16	and	and	CCONJ
fcis-9479	83	17	the	the	DET
fcis-9479	83	18	number	number	NOUN
fcis-9479	83	19	of	of	ADP
fcis-9479	83	20	features	feature	NOUN
fcis-9479	83	21	selected	select	VERB
fcis-9479	83	22	was	be	AUX
fcis-9479	83	23	299	299	NUM
fcis-9479	83	24	.	.	PUNCT
fcis-9479	84	1	the	the	DET
fcis-9479	84	2	minimum	minimum	ADJ
fcis-9479	84	3	loss	loss	NOUN
fcis-9479	84	4	value	value	NOUN
fcis-9479	84	5	obtained	obtain	VERB
fcis-9479	84	6	during	during	ADP
fcis-9479	84	7	training	training	NOUN
fcis-9479	84	8	was	be	AUX
fcis-9479	84	9	0.193351	0.193351	NUM
fcis-9479	84	10	and	and	CCONJ
fcis-9479	84	11	the	the	DET
fcis-9479	84	12	highest	high	ADJ
fcis-9479	84	13	accuracy	accuracy	NOUN
fcis-9479	84	14	value	value	NOUN
fcis-9479	84	15	obtained	obtain	VERB
fcis-9479	84	16	on	on	ADP
fcis-9479	84	17	the	the	DET
fcis-9479	84	18	validation	validation	NOUN
fcis-9479	84	19	set	set	NOUN
fcis-9479	84	20	was	be	AUX
fcis-9479	84	21	0.976451	0.976451	NUM
fcis-9479	84	22	.	.	PUNCT
fcis-9479	85	1	the	the	DET
fcis-9479	85	2	xgboost	xgboost	PROPN
fcis-9479	85	3	model	model	PROPN
fcis-9479	85	4	training	training	NOUN
fcis-9479	85	5	process	process	NOUN
fcis-9479	85	6	is	be	AUX
fcis-9479	85	7	shown	show	VERB
fcis-9479	85	8	in	in	ADP
fcis-9479	85	9	figure	figure	NOUN
fcis-9479	85	10	4	4	NUM
fcis-9479	85	11	,	,	PUNCT
fcis-9479	85	12	figure	figure	VERB
fcis-9479	85	13	5	5	NUM
fcis-9479	85	14	and	and	CCONJ
fcis-9479	85	15	figure	figure	VERB
fcis-9479	85	16	6	6	NUM
fcis-9479	85	17	.	.	PUNCT
fcis-9479	86	1	it	it	PRON
fcis-9479	86	2	can	can	AUX
fcis-9479	86	3	be	be	AUX
fcis-9479	86	4	seen	see	VERB
fcis-9479	86	5	from	from	ADP
fcis-9479	86	6	figure	figure	NOUN
fcis-9479	86	7	4	4	NUM
fcis-9479	86	8	that	that	SCONJ
fcis-9479	86	9	31160	31160	NUM
fcis-9479	86	10	pieces	piece	NOUN
fcis-9479	86	11	of	of	ADP
fcis-9479	86	12	data	datum	NOUN
fcis-9479	86	13	were	be	AUX
fcis-9479	86	14	used	use	VERB
fcis-9479	86	15	for	for	ADP
fcis-9479	86	16	training	training	NOUN
fcis-9479	86	17	during	during	ADP
fcis-9479	86	18	this	this	DET
fcis-9479	86	19	training	training	NOUN
fcis-9479	86	20	process	process	NOUN
fcis-9479	86	21	.	.	PUNCT
fcis-9479	87	1	during	during	ADP
fcis-9479	87	2	the	the	DET
fcis-9479	87	3	training	training	NOUN
fcis-9479	87	4	process	process	NOUN
fcis-9479	87	5	,	,	PUNCT
fcis-9479	87	6	it	it	PRON
fcis-9479	87	7	can	can	AUX
fcis-9479	87	8	be	be	AUX
fcis-9479	87	9	seen	see	VERB
fcis-9479	87	10	from	from	ADP
fcis-9479	87	11	figure	figure	NOUN
fcis-9479	87	12	5	5	NUM
fcis-9479	87	13	that	that	SCONJ
fcis-9479	87	14	the	the	DET
fcis-9479	87	15	highest	high	ADJ
fcis-9479	87	16	training	training	NOUN
fcis-9479	87	17	accuracy	accuracy	NOUN
fcis-9479	87	18	of	of	ADP
fcis-9479	87	19	the	the	DET
fcis-9479	87	20	model	model	NOUN
fcis-9479	87	21	during	during	ADP
fcis-9479	87	22	training	training	NOUN
fcis-9479	87	23	could	could	AUX
fcis-9479	87	24	reach	reach	VERB
fcis-9479	87	25	0.99981	0.99981	NUM
fcis-9479	87	26	.	.	PUNCT
fcis-9479	88	1	it	it	PRON
fcis-9479	88	2	can	can	AUX
fcis-9479	88	3	be	be	AUX
fcis-9479	88	4	seen	see	VERB
fcis-9479	88	5	from	from	ADP
fcis-9479	88	6	figure	figure	NOUN
fcis-9479	88	7	6	6	NUM
fcis-9479	88	8	that	that	SCONJ
fcis-9479	88	9	through	through	ADP
fcis-9479	88	10	multiple	multiple	ADJ
fcis-9479	88	11	iterations	iteration	NOUN
fcis-9479	88	12	,	,	PUNCT
fcis-9479	88	13	the	the	DET
fcis-9479	88	14	highest	high	ADJ
fcis-9479	88	15	f1	f1	ADJ
fcis-9479	88	16	score	score	NOUN
fcis-9479	88	17	of	of	ADP
fcis-9479	88	18	xgboost	xgboost	ADV
fcis-9479	88	19	on	on	ADP
fcis-9479	88	20	the	the	DET
fcis-9479	88	21	validation	validation	NOUN
fcis-9479	88	22	set	set	NOUN
fcis-9479	88	23	could	could	AUX
fcis-9479	88	24	reach	reach	VERB
fcis-9479	88	25	0.96589	0.96589	NUM
fcis-9479	88	26	.	.	PUNCT
fcis-9479	89	1	figure	figure	NOUN
fcis-9479	89	2	3	3	NUM
fcis-9479	89	3	.	.	NOUN
fcis-9479	89	4	lightgbm	lightgbm	ADJ
fcis-9479	89	5	model	model	NOUN
fcis-9479	89	6	training	training	NOUN
fcis-9479	89	7	process	process	NOUN
fcis-9479	89	8	figure	figure	NOUN
fcis-9479	89	9	4	4	NUM
fcis-9479	89	10	.	.	PUNCT
fcis-9479	90	1	xgboost	xgboost	PROPN
fcis-9479	91	1	model	model	NOUN
fcis-9479	91	2	training	training	NOUN
fcis-9479	91	3	process	process	NOUN
fcis-9479	91	4	figure	figure	NOUN
fcis-9479	91	5	5	5	NUM
fcis-9479	91	6	.	.	PUNCT
fcis-9479	92	1	xgboost	xgboost	ADJ
fcis-9479	92	2	training	training	NOUN
fcis-9479	92	3	accuracy	accuracy	NOUN
fcis-9479	92	4	figure	figure	NOUN
fcis-9479	92	5	6	6	NUM
fcis-9479	92	6	.	.	PUNCT
fcis-9479	93	1	xgboost	xgboost	VERB
fcis-9479	93	2	training	training	NOUN
fcis-9479	93	3	f1	f1	PROPN
fcis-9479	93	4	value	value	NOUN
fcis-9479	93	5	4.2.2	4.2.2	X
fcis-9479	93	6	.	.	PUNCT
fcis-9479	93	7	model	model	NOUN
fcis-9479	93	8	testing	testing	NOUN
fcis-9479	93	9	process	process	NOUN
fcis-9479	93	10	figure	figure	VERB
fcis-9479	93	11	7	7	NUM
fcis-9479	93	12	.	.	PUNCT
fcis-9479	94	1	the	the	DET
fcis-9479	94	2	test	test	NOUN
fcis-9479	94	3	accuracy	accuracy	NOUN
fcis-9479	94	4	of	of	ADP
fcis-9479	94	5	lightgbm	lightgbm	ADJ
fcis-9479	94	6	108	108	NUM
fcis-9479	94	7	the	the	DET
fcis-9479	94	8	lightgbm	lightgbm	ADJ
fcis-9479	94	9	test	test	NOUN
fcis-9479	94	10	process	process	NOUN
fcis-9479	94	11	is	be	AUX
fcis-9479	94	12	shown	show	VERB
fcis-9479	94	13	in	in	ADP
fcis-9479	94	14	figure	figure	NOUN
fcis-9479	94	15	7	7	NUM
fcis-9479	94	16	.	.	PUNCT
fcis-9479	94	17	during	during	ADP
fcis-9479	94	18	the	the	DET
fcis-9479	94	19	test	test	NOUN
fcis-9479	94	20	,	,	PUNCT
fcis-9479	94	21	4000	4000	NUM
fcis-9479	94	22	pieces	piece	NOUN
fcis-9479	94	23	of	of	ADP
fcis-9479	94	24	data	datum	NOUN
fcis-9479	94	25	were	be	AUX
fcis-9479	94	26	selected	select	VERB
fcis-9479	94	27	for	for	ADP
fcis-9479	94	28	testing	testing	NOUN
fcis-9479	94	29	,	,	PUNCT
fcis-9479	94	30	and	and	CCONJ
fcis-9479	94	31	it	it	PRON
fcis-9479	94	32	can	can	AUX
fcis-9479	94	33	be	be	AUX
fcis-9479	94	34	seen	see	VERB
fcis-9479	94	35	from	from	ADP
fcis-9479	94	36	the	the	DET
fcis-9479	94	37	figure7	figure7	NOUN
fcis-9479	94	38	that	that	SCONJ
fcis-9479	94	39	the	the	DET
fcis-9479	94	40	test	test	NOUN
fcis-9479	94	41	accuracy	accuracy	NOUN
fcis-9479	94	42	reached	reach	VERB
fcis-9479	94	43	up	up	ADP
fcis-9479	94	44	to	to	ADP
fcis-9479	94	45	0.99542	0.99542	NUM
fcis-9479	94	46	.	.	PUNCT
fcis-9479	95	1	4.2.3	4.2.3	X
fcis-9479	95	2	.	.	PUNCT
fcis-9479	96	1	comparison	comparison	NOUN
fcis-9479	96	2	of	of	ADP
fcis-9479	96	3	experimental	experimental	ADJ
fcis-9479	96	4	results	result	NOUN
fcis-9479	96	5	the	the	DET
fcis-9479	96	6	main	main	ADJ
fcis-9479	96	7	performance	performance	NOUN
fcis-9479	96	8	indicators	indicator	NOUN
fcis-9479	96	9	of	of	ADP
fcis-9479	96	10	the	the	DET
fcis-9479	96	11	evaluation	evaluation	NOUN
fcis-9479	96	12	model	model	NOUN
fcis-9479	96	13	are	be	AUX
fcis-9479	96	14	:	:	PUNCT
fcis-9479	96	15	100	100	NUM
fcis-9479	96	16	%	%	NOUN
fcis-9479	96	17	(	(	PUNCT
fcis-9479	96	18	1	1	NUM
fcis-9479	96	19	)	)	PUNCT
fcis-9479	96	20	where	where	SCONJ
fcis-9479	96	21	tp	tp	PART
fcis-9479	96	22	denotes	denote	VERB
fcis-9479	96	23	true	true	ADJ
fcis-9479	96	24	cases	case	NOUN
fcis-9479	96	25	,	,	PUNCT
fcis-9479	96	26	fp	fp	PROPN
fcis-9479	96	27	denotes	denote	NOUN
fcis-9479	96	28	pseudo	pseudo	ADJ
fcis-9479	96	29	-	-	ADJ
fcis-9479	96	30	positive	positive	ADJ
fcis-9479	96	31	cases	case	NOUN
fcis-9479	96	32	,	,	PUNCT
fcis-9479	96	33	and	and	CCONJ
fcis-9479	96	34	precision	precision	NOUN
fcis-9479	96	35	denotes	denote	VERB
fcis-9479	96	36	the	the	DET
fcis-9479	96	37	proportion	proportion	NOUN
fcis-9479	96	38	of	of	ADP
fcis-9479	96	39	samples	sample	NOUN
fcis-9479	96	40	predicted	predict	VERB
fcis-9479	96	41	to	to	PART
fcis-9479	96	42	be	be	AUX
fcis-9479	96	43	true	true	ADJ
fcis-9479	96	44	positive	positive	ADJ
fcis-9479	96	45	cases	case	NOUN
fcis-9479	96	46	in	in	ADP
fcis-9479	96	47	the	the	DET
fcis-9479	96	48	sample	sample	NOUN
fcis-9479	96	49	of	of	ADP
fcis-9479	96	50	positive	positive	ADJ
fcis-9479	96	51	cases	case	NOUN
fcis-9479	96	52	.	.	PUNCT
fcis-9479	97	1	100	100	NUM
fcis-9479	97	2	%	%	NOUN
fcis-9479	97	3	(	(	PUNCT
fcis-9479	97	4	2	2	NUM
fcis-9479	97	5	)	)	PUNCT
fcis-9479	97	6	where	where	SCONJ
fcis-9479	97	7	tp	tp	PART
fcis-9479	97	8	denotes	denote	VERB
fcis-9479	97	9	true	true	ADJ
fcis-9479	97	10	cases	case	NOUN
fcis-9479	97	11	,	,	PUNCT
fcis-9479	97	12	fn	fn	NOUN
fcis-9479	97	13	denotes	denote	NOUN
fcis-9479	97	14	pseudo	pseudo	NOUN
fcis-9479	97	15	counter	counter	NOUN
fcis-9479	97	16	examples	example	NOUN
fcis-9479	97	17	,	,	PUNCT
fcis-9479	97	18	and	and	CCONJ
fcis-9479	97	19	recall	recall	NOUN
fcis-9479	97	20	denotes	denote	VERB
fcis-9479	97	21	the	the	DET
fcis-9479	97	22	proportion	proportion	NOUN
fcis-9479	97	23	of	of	ADP
fcis-9479	97	24	samples	sample	NOUN
fcis-9479	97	25	predicted	predict	VERB
fcis-9479	97	26	as	as	ADP
fcis-9479	97	27	positive	positive	ADJ
fcis-9479	97	28	cases	case	NOUN
fcis-9479	97	29	to	to	ADP
fcis-9479	97	30	all	all	DET
fcis-9479	97	31	positive	positive	ADJ
fcis-9479	97	32	samples	sample	NOUN
fcis-9479	97	33	.	.	PUNCT
fcis-9479	98	1	1	1	NUM
fcis-9479	98	2	(	(	PUNCT
fcis-9479	98	3	3	3	NUM
fcis-9479	98	4	)	)	PUNCT
fcis-9479	98	5	the	the	DET
fcis-9479	98	6	f1	f1	PROPN
fcis-9479	98	7	values	value	NOUN
fcis-9479	98	8	obtained	obtain	VERB
fcis-9479	98	9	by	by	ADP
fcis-9479	98	10	different	different	ADJ
fcis-9479	98	11	models	model	NOUN
fcis-9479	98	12	after	after	SCONJ
fcis-9479	98	13	training	training	NOUN
fcis-9479	98	14	are	be	AUX
fcis-9479	98	15	shown	show	VERB
fcis-9479	98	16	in	in	ADP
fcis-9479	98	17	table	table	NOUN
fcis-9479	98	18	4	4	NUM
fcis-9479	98	19	.	.	PUNCT
fcis-9479	99	1	after	after	ADP
fcis-9479	99	2	comparing	compare	VERB
fcis-9479	99	3	different	different	ADJ
fcis-9479	99	4	pre	pre	ADJ
fcis-9479	99	5	-	-	ADJ
fcis-9479	99	6	trained	train	VERB
fcis-9479	99	7	language	language	NOUN
fcis-9479	99	8	models	model	NOUN
fcis-9479	99	9	under	under	ADP
fcis-9479	99	10	the	the	DET
fcis-9479	99	11	same	same	ADJ
fcis-9479	99	12	classification	classification	NOUN
fcis-9479	99	13	algorithm	algorithm	NOUN
fcis-9479	99	14	,	,	PUNCT
fcis-9479	99	15	the	the	DET
fcis-9479	99	16	prediction	prediction	NOUN
fcis-9479	99	17	result	result	NOUN
fcis-9479	99	18	of	of	ADP
fcis-9479	99	19	the	the	DET
fcis-9479	99	20	xgboost	xgboost	PROPN
fcis-9479	99	21	model	model	NOUN
fcis-9479	99	22	is	be	AUX
fcis-9479	99	23	significantly	significantly	ADV
fcis-9479	99	24	higher	high	ADJ
fcis-9479	99	25	than	than	ADP
fcis-9479	99	26	that	that	PRON
fcis-9479	99	27	of	of	ADP
fcis-9479	99	28	lightgbm	lightgbm	ADJ
fcis-9479	99	29	,	,	PUNCT
fcis-9479	99	30	and	and	CCONJ
fcis-9479	99	31	it	it	PRON
fcis-9479	99	32	is	be	AUX
fcis-9479	99	33	also	also	ADV
fcis-9479	99	34	higher	high	ADJ
fcis-9479	99	35	than	than	ADP
fcis-9479	99	36	the	the	DET
fcis-9479	99	37	result	result	NOUN
fcis-9479	99	38	of	of	ADP
fcis-9479	99	39	the	the	DET
fcis-9479	99	40	fusion	fusion	NOUN
fcis-9479	99	41	of	of	ADP
fcis-9479	99	42	cnn	cnn	PROPN
fcis-9479	99	43	and	and	CCONJ
fcis-9479	99	44	lstm	lstm	NOUN
fcis-9479	99	45	models	model	NOUN
fcis-9479	99	46	.	.	PUNCT
fcis-9479	100	1	the	the	DET
fcis-9479	100	2	analysis	analysis	NOUN
fcis-9479	100	3	shows	show	VERB
fcis-9479	100	4	that	that	SCONJ
fcis-9479	100	5	the	the	DET
fcis-9479	100	6	xgboost	xgboost	PROPN
fcis-9479	100	7	model	model	NOUN
fcis-9479	100	8	is	be	AUX
fcis-9479	100	9	more	more	ADV
fcis-9479	100	10	advantageous	advantageous	ADJ
fcis-9479	100	11	,	,	PUNCT
fcis-9479	100	12	and	and	CCONJ
fcis-9479	100	13	the	the	DET
fcis-9479	100	14	f1	f1	NOUN
fcis-9479	100	15	value	value	NOUN
fcis-9479	100	16	reaches	reach	VERB
fcis-9479	100	17	0.972	0.972	NUM
fcis-9479	100	18	.	.	PUNCT
fcis-9479	100	19	table	table	NOUN
fcis-9479	100	20	4	4	NUM
fcis-9479	100	21	.	.	PUNCT
fcis-9479	100	22	experimental	experimental	ADJ
fcis-9479	100	23	results	result	NOUN
fcis-9479	100	24	model	model	PROPN
fcis-9479	100	25	f1	f1	PROPN
fcis-9479	100	26	score	score	NOUN
fcis-9479	100	27	cnn	cnn	PROPN
fcis-9479	100	28	and	and	CCONJ
fcis-9479	100	29	lstm	lstm	PROPN
fcis-9479	100	30	model	model	NOUN
fcis-9479	100	31	fusion	fusion	NOUN
fcis-9479	100	32	0.917	0.917	NUM
fcis-9479	100	33	lightgbm	lightgbm	VERB
fcis-9479	100	34	0.918	0.918	NUM
fcis-9479	100	35	xgboost	xgboost	ADV
fcis-9479	100	36	0.972	0.972	NUM
fcis-9479	100	37	5	5	NUM
fcis-9479	100	38	.	.	PUNCT
fcis-9479	101	1	conclusion	conclusion	NOUN
fcis-9479	101	2	this	this	DET
fcis-9479	101	3	paper	paper	NOUN
fcis-9479	101	4	compares	compare	VERB
fcis-9479	101	5	the	the	DET
fcis-9479	101	6	detection	detection	NOUN
fcis-9479	101	7	results	result	NOUN
fcis-9479	101	8	of	of	ADP
fcis-9479	101	9	cnn	cnn	PROPN
fcis-9479	101	10	and	and	CCONJ
fcis-9479	101	11	lstm	lstm	PROPN
fcis-9479	101	12	model	model	NOUN
fcis-9479	101	13	fusion	fusion	NOUN
fcis-9479	101	14	,	,	PUNCT
fcis-9479	101	15	lightgbm	lightgbm	ADJ
fcis-9479	101	16	,	,	PUNCT
fcis-9479	101	17	and	and	CCONJ
fcis-9479	101	18	xgboost	xgboost	ADJ
fcis-9479	101	19	models	model	NOUN
fcis-9479	101	20	.	.	PUNCT
fcis-9479	102	1	the	the	DET
fcis-9479	102	2	experimental	experimental	ADJ
fcis-9479	102	3	results	result	NOUN
fcis-9479	102	4	show	show	VERB
fcis-9479	102	5	that	that	SCONJ
fcis-9479	102	6	xgboost	xgboost	PROPN
fcis-9479	102	7	outperforms	outperform	VERB
fcis-9479	102	8	the	the	DET
fcis-9479	102	9	other	other	ADJ
fcis-9479	102	10	models	model	NOUN
fcis-9479	102	11	in	in	ADP
fcis-9479	102	12	this	this	DET
fcis-9479	102	13	study	study	NOUN
fcis-9479	102	14	in	in	ADP
fcis-9479	102	15	terms	term	NOUN
fcis-9479	102	16	of	of	ADP
fcis-9479	102	17	performance	performance	NOUN
fcis-9479	102	18	metrics	metric	NOUN
fcis-9479	102	19	,	,	PUNCT
fcis-9479	102	20	and	and	CCONJ
fcis-9479	102	21	its	its	PRON
fcis-9479	102	22	f1	f1	NOUN
fcis-9479	102	23	values	value	NOUN
fcis-9479	102	24	are	be	AUX
fcis-9479	102	25	all	all	ADV
fcis-9479	102	26	higher	high	ADJ
fcis-9479	102	27	than	than	ADP
fcis-9479	102	28	the	the	DET
fcis-9479	102	29	other	other	ADJ
fcis-9479	102	30	models	model	NOUN
fcis-9479	102	31	.	.	PUNCT
fcis-9479	103	1	of	of	ADP
fcis-9479	103	2	course	course	NOUN
fcis-9479	103	3	,	,	PUNCT
fcis-9479	103	4	there	there	PRON
fcis-9479	103	5	are	be	VERB
fcis-9479	103	6	still	still	ADV
fcis-9479	103	7	shortcomings	shortcoming	NOUN
fcis-9479	103	8	in	in	ADP
fcis-9479	103	9	this	this	DET
fcis-9479	103	10	paper	paper	NOUN
fcis-9479	103	11	for	for	ADP
fcis-9479	103	12	the	the	DET
fcis-9479	103	13	study	study	NOUN
fcis-9479	103	14	of	of	ADP
fcis-9479	103	15	fake	fake	ADJ
fcis-9479	103	16	news	news	NOUN
fcis-9479	103	17	texts	text	NOUN
fcis-9479	103	18	.	.	PUNCT
fcis-9479	104	1	longitudinal	longitudinal	ADJ
fcis-9479	104	2	analysis	analysis	NOUN
fcis-9479	104	3	and	and	CCONJ
fcis-9479	104	4	comparison	comparison	NOUN
fcis-9479	104	5	from	from	ADP
fcis-9479	104	6	multiple	multiple	ADJ
fcis-9479	104	7	classification	classification	NOUN
fcis-9479	104	8	algorithms	algorithm	NOUN
fcis-9479	104	9	are	be	AUX
fcis-9479	104	10	still	still	ADV
fcis-9479	104	11	needed	need	VERB
fcis-9479	104	12	to	to	PART
fcis-9479	104	13	find	find	VERB
fcis-9479	104	14	the	the	DET
fcis-9479	104	15	best	good	ADJ
fcis-9479	104	16	classification	classification	NOUN
fcis-9479	104	17	algorithm	algorithm	NOUN
fcis-9479	104	18	.	.	PUNCT
fcis-9479	105	1	in	in	ADP
fcis-9479	105	2	this	this	DET
fcis-9479	105	3	paper	paper	NOUN
fcis-9479	105	4	,	,	PUNCT
fcis-9479	105	5	only	only	ADV
fcis-9479	105	6	single	single	ADJ
fcis-9479	105	7	-	-	PUNCT
fcis-9479	105	8	mode	mode	NOUN
fcis-9479	105	9	false	false	ADJ
fcis-9479	105	10	news	news	NOUN
fcis-9479	105	11	was	be	AUX
fcis-9479	105	12	detected	detect	VERB
fcis-9479	105	13	in	in	ADP
fcis-9479	105	14	the	the	DET
fcis-9479	105	15	experiment	experiment	NOUN
fcis-9479	105	16	.	.	PUNCT
fcis-9479	106	1	in	in	ADP
fcis-9479	106	2	the	the	DET
fcis-9479	106	3	next	next	ADJ
fcis-9479	106	4	step	step	NOUN
fcis-9479	106	5	,	,	PUNCT
fcis-9479	106	6	the	the	DET
fcis-9479	106	7	method	method	NOUN
fcis-9479	106	8	proposed	propose	VERB
fcis-9479	106	9	in	in	ADP
fcis-9479	106	10	this	this	DET
fcis-9479	106	11	paper	paper	NOUN
fcis-9479	106	12	will	will	AUX
fcis-9479	106	13	be	be	AUX
fcis-9479	106	14	used	use	VERB
fcis-9479	106	15	to	to	PART
fcis-9479	106	16	conduct	conduct	VERB
fcis-9479	106	17	experiments	experiment	NOUN
fcis-9479	106	18	on	on	ADP
fcis-9479	106	19	multi	multi	ADJ
fcis-9479	106	20	-	-	ADJ
fcis-9479	106	21	mode	mode	ADJ
fcis-9479	106	22	false	false	ADJ
fcis-9479	106	23	news	news	NOUN
fcis-9479	106	24	.	.	PUNCT
fcis-9479	107	1	references	reference	NOUN
fcis-9479	107	2	[	[	X
fcis-9479	107	3	1	1	X
fcis-9479	107	4	]	]	PUNCT
fcis-9479	107	5	yujun	yujun	PROPN
fcis-9479	107	6	gao	gao	PROPN
fcis-9479	107	7	,	,	PUNCT
fcis-9479	107	8	gang	gang	PROPN
fcis-9479	107	9	liang	liang	PROPN
fcis-9479	107	10	,	,	PUNCT
fcis-9479	107	11	fangting	fangting	PROPN
fcis-9479	107	12	jiang	jiang	PROPN
fcis-9479	107	13	,	,	PUNCT
fcis-9479	107	14	chun	chun	PROPN
fcis-9479	107	15	xu	xu	PROPN
fcis-9479	107	16	,	,	PUNCT
fcis-9479	107	17	jin	jin	PROPN
fcis-9479	107	18	yang	yang	PROPN
fcis-9479	107	19	,	,	PUNCT
fcis-9479	107	20	junren	junren	PROPN
fcis-9479	107	21	chen	chen	PROPN
fcis-9479	107	22	,	,	PUNCT
fcis-9479	107	23	hao	hao	PROPN
fcis-9479	107	24	wang.a	wang.a	PROPN
fcis-9479	107	25	review	review	NOUN
fcis-9479	107	26	of	of	ADP
fcis-9479	107	27	social	social	ADJ
fcis-9479	107	28	network	network	NOUN
fcis-9479	107	29	humor	humor	NOUN
fcis-9479	107	30	detection[j	detection[j	PROPN
fcis-9479	107	31	]	]	PUNCT
fcis-9479	107	32	.	.	PUNCT
fcis-9479	108	1	journal	journal	PROPN
fcis-9479	108	2	of	of	ADP
fcis-9479	108	3	electronics	electronic	NOUN
fcis-9479	108	4	,	,	PUNCT
fcis-9479	108	5	2020	2020	NUM
fcis-9479	108	6	,	,	PUNCT
fcis-9479	108	7	48(7	48(7	NUM
fcis-9479	108	8	):	):	PUNCT
fcis-9479	108	9	1421	1421	NUM
fcis-9479	108	10	-	-	SYM
fcis-9479	108	11	1435	1435	NUM
fcis-9479	108	12	.	.	PUNCT
fcis-9479	109	1	[	[	X
fcis-9479	109	2	2	2	X
fcis-9479	109	3	]	]	PUNCT
fcis-9479	109	4	shengsheng	shengsheng	PROPN
fcis-9479	109	5	qian	qian	PROPN
fcis-9479	109	6	,	,	PUNCT
fcis-9479	109	7	tianzhu	tianzhu	PROPN
fcis-9479	109	8	zhanng	zhanng	PROPN
fcis-9479	109	9	,	,	PUNCT
fcis-9479	109	10	changsheng	changsheng	PROPN
fcis-9479	109	11	xu.a	xu.a	PROPN
fcis-9479	109	12	review	review	PROPN
fcis-9479	109	13	of	of	ADP
fcis-9479	109	14	multimedia	multimedia	ADJ
fcis-9479	109	15	social	social	ADJ
fcis-9479	109	16	event	event	NOUN
fcis-9479	109	17	analysis[j].computer	analysis[j].computer	PROPN
fcis-9479	109	18	science	science	NOUN
fcis-9479	109	19	,	,	PUNCT
fcis-9479	109	20	2021	2021	NUM
fcis-9479	109	21	,	,	PUNCT
fcis-9479	109	22	48(3	48(3	NUM
fcis-9479	109	23	):	):	PUNCT
fcis-9479	109	24	97	97	NUM
fcis-9479	109	25	-	-	SYM
fcis-9479	109	26	112	112	NUM
fcis-9479	109	27	.	.	PUNCT
fcis-9479	110	1	[	[	X
fcis-9479	110	2	3	3	X
fcis-9479	110	3	]	]	X
fcis-9479	110	4	chen	chen	PROPN
fcis-9479	110	5	l	l	PROPN
fcis-9479	110	6	,	,	PUNCT
fcis-9479	110	7	liang	liang	PROPN
fcis-9479	110	8	j	j	PROPN
fcis-9479	110	9	,	,	PUNCT
fcis-9479	110	10	xie	xie	PROPN
fcis-9479	110	11	c	c	PROPN
fcis-9479	110	12	,	,	PUNCT
fcis-9479	110	13	xiao	xiao	PROPN
fcis-9479	110	14	y.	y.	PROPN
fcis-9479	110	15	short	short	ADJ
fcis-9479	110	16	text	text	NOUN
fcis-9479	110	17	entity	entity	NOUN
fcis-9479	110	18	linking	link	VERB
fcis-9479	110	19	with	with	ADP
fcis-9479	110	20	fine	fine	ADV
fcis-9479	110	21	-	-	PUNCT
fcis-9479	110	22	grained	grain	VERB
fcis-9479	110	23	topics[c]//proceedings	topics[c]//proceeding	NOUN
fcis-9479	110	24	of	of	ADP
fcis-9479	110	25	the	the	DET
fcis-9479	110	26	27th	27th	ADJ
fcis-9479	110	27	acm	acm	PROPN
fcis-9479	110	28	international	international	ADJ
fcis-9479	110	29	conference	conference	NOUN
fcis-9479	110	30	on	on	ADP
fcis-9479	110	31	information	information	NOUN
fcis-9479	110	32	and	and	CCONJ
fcis-9479	110	33	knowledge	knowledge	NOUN
fcis-9479	110	34	management	management	NOUN
fcis-9479	110	35	.	.	PUNCT
fcis-9479	111	1	torino	torino	PROPN
fcis-9479	111	2	,	,	PUNCT
fcis-9479	111	3	italy	italy	PROPN
fcis-9479	111	4	:	:	PUNCT
fcis-9479	111	5	association	association	NOUN
fcis-9479	111	6	for	for	ADP
fcis-9479	111	7	computing	compute	VERB
fcis-9479	111	8	machinery	machinery	NOUN
fcis-9479	111	9	,	,	PUNCT
fcis-9479	111	10	2018	2018	NUM
fcis-9479	111	11	:	:	PUNCT
fcis-9479	111	12	457	457	NUM
fcis-9479	111	13	-	-	SYM
fcis-9479	111	14	466	466	NUM
fcis-9479	111	15	.	.	PUNCT
fcis-9479	112	1	[	[	X
fcis-9479	112	2	4	4	NUM
fcis-9479	112	3	]	]	PUNCT
fcis-9479	112	4	yu	yu	PROPN
fcis-9479	112	5	f	f	PROPN
fcis-9479	112	6	,	,	PUNCT
fcis-9479	112	7	liu	liu	PROPN
fcis-9479	112	8	q	q	PROPN
fcis-9479	112	9	,	,	PUNCT
fcis-9479	112	10	wu	wu	PROPN
fcis-9479	112	11	s	s	PROPN
fcis-9479	112	12	,	,	PUNCT
fcis-9479	112	13	wang	wang	PROPN
fcis-9479	112	14	l	l	PROPN
fcis-9479	112	15	,	,	PUNCT
fcis-9479	112	16	tan	tan	PROPN
fcis-9479	112	17	t.	t.	VERB
fcis-9479	112	18	a	a	DET
fcis-9479	112	19	convolutional	convolutional	ADJ
fcis-9479	112	20	approach	approach	NOUN
fcis-9479	112	21	for	for	ADP
fcis-9479	112	22	misinformation	misinformation	NOUN
fcis-9479	112	23	identification	identification	NOUN
fcis-9479	112	24	.	.	PUNCT
fcis-9479	113	1	[	[	X
fcis-9479	113	2	c]//ijcai	c]//ijcai	PROPN
fcis-9479	113	3	.	.	PROPN
fcis-9479	113	4	2017	2017	NUM
fcis-9479	113	5	:	:	PUNCT
fcis-9479	113	6	3901	3901	NUM
fcis-9479	113	7	-	-	SYM
fcis-9479	113	8	3907	3907	NUM
fcis-9479	113	9	.	.	PUNCT
fcis-9479	114	1	[	[	X
fcis-9479	114	2	5	5	NUM
fcis-9479	114	3	]	]	PUNCT
fcis-9479	114	4	ma	ma	PROPN
fcis-9479	114	5	j	j	PROPN
fcis-9479	114	6	,	,	PUNCT
fcis-9479	114	7	gao	gao	PROPN
fcis-9479	114	8	w	w	PROPN
fcis-9479	114	9	,	,	PUNCT
fcis-9479	114	10	wei	wei	PROPN
fcis-9479	114	11	z	z	PROPN
fcis-9479	114	12	,	,	PUNCT
fcis-9479	114	13	lu	lu	PROPN
fcis-9479	114	14	y	y	PROPN
fcis-9479	114	15	,	,	PUNCT
fcis-9479	114	16	wong	wong	PROPN
fcis-9479	114	17	k	k	PROPN
fcis-9479	114	18	f.	f.	PROPN
fcis-9479	114	19	detect	detect	VERB
fcis-9479	114	20	rumors	rumor	NOUN
fcis-9479	114	21	using	use	VERB
fcis-9479	114	22	time	time	NOUN
fcis-9479	114	23	series	series	NOUN
fcis-9479	114	24	of	of	ADP
fcis-9479	114	25	social	social	ADJ
fcis-9479	114	26	context	context	NOUN
fcis-9479	114	27	information	information	NOUN
fcis-9479	114	28	on	on	ADP
fcis-9479	114	29	microblogging	microblogge	VERB
fcis-9479	114	30	[	[	X
fcis-9479	114	31	m]//social	m]//social	ADJ
fcis-9479	114	32	media	medium	NOUN
fcis-9479	114	33	content	content	NOUN
fcis-9479	114	34	analysis	analysis	NOUN
fcis-9479	114	35	:	:	PUNCT
fcis-9479	114	36	natural	natural	ADJ
fcis-9479	114	37	language	language	NOUN
fcis-9479	114	38	processing	processing	NOUN
fcis-9479	114	39	and	and	CCONJ
fcis-9479	114	40	beyond	beyond	ADP
fcis-9479	114	41	.	.	PUNCT
fcis-9479	115	1	world	world	NOUN
fcis-9479	115	2	scientific	scientific	ADJ
fcis-9479	115	3	,	,	PUNCT
fcis-9479	115	4	2018	2018	NUM
fcis-9479	115	5	:	:	PUNCT
fcis-9479	115	6	6777	6777	NUM
fcis-9479	115	7	.	.	PUNCT
fcis-9479	116	1	[	[	X
fcis-9479	116	2	6	6	NUM
fcis-9479	116	3	]	]	PUNCT
fcis-9479	116	4	qi	qi	PROPN
fcis-9479	116	5	p	p	PROPN
fcis-9479	116	6	,	,	PUNCT
fcis-9479	116	7	cao	cao	PROPN
fcis-9479	116	8	j	j	PROPN
fcis-9479	116	9	,	,	PUNCT
fcis-9479	116	10	yang	yang	PROPN
fcis-9479	116	11	t	t	PROPN
fcis-9479	116	12	,	,	PUNCT
fcis-9479	116	13	guo	guo	PROPN
fcis-9479	116	14	j	j	PROPN
fcis-9479	116	15	,	,	PUNCT
fcis-9479	116	16	li	li	PROPN
fcis-9479	116	17	j.	j.	PROPN
fcis-9479	116	18	exploiting	exploit	VERB
fcis-9479	116	19	multi	multi	ADJ
fcis-9479	116	20	-	-	ADJ
fcis-9479	116	21	domain	domain	ADJ
fcis-9479	116	22	visual	visual	ADJ
fcis-9479	116	23	information	information	NOUN
fcis-9479	116	24	for	for	ADP
fcis-9479	116	25	fake	fake	ADJ
fcis-9479	116	26	news	news	NOUN
fcis-9479	116	27	detection[c]//2019	detection[c]//2019	PUNCT
fcis-9479	116	28	ieee	ieee	PROPN
fcis-9479	116	29	international	international	ADJ
fcis-9479	116	30	conference	conference	NOUN
fcis-9479	116	31	on	on	ADP
fcis-9479	116	32	data	datum	NOUN
fcis-9479	116	33	mining	mining	NOUN
fcis-9479	116	34	(	(	PUNCT
fcis-9479	116	35	icdm	icdm	NOUN
fcis-9479	116	36	)	)	PUNCT
fcis-9479	116	37	.	.	PUNCT
fcis-9479	117	1	ieee	ieee	PROPN
fcis-9479	117	2	,	,	PUNCT
fcis-9479	117	3	2019	2019	NUM
fcis-9479	117	4	:	:	PUNCT
fcis-9479	117	5	518	518	NUM
fcis-9479	117	6	-	-	SYM
fcis-9479	117	7	527	527	NUM
fcis-9479	117	8	.	.	PUNCT
fcis-9479	118	1	[	[	X
fcis-9479	118	2	7	7	X
fcis-9479	118	3	]	]	X
fcis-9479	118	4	gupta	gupta	PROPN
fcis-9479	118	5	m	m	PROPN
fcis-9479	118	6	,	,	PUNCT
fcis-9479	118	7	zhao	zhao	PROPN
fcis-9479	118	8	p	p	PROPN
fcis-9479	118	9	,	,	PUNCT
fcis-9479	118	10	han	han	PROPN
fcis-9479	118	11	j.	j.	PROPN
fcis-9479	118	12	evaluating	evaluating	PROPN
fcis-9479	118	13	event	event	NOUN
fcis-9479	118	14	credibility	credibility	NOUN
fcis-9479	118	15	on	on	ADP
fcis-9479	118	16	twitter	twitter	NOUN
fcis-9479	118	17	[	[	X
fcis-9479	118	18	c]//proceedings	c]//proceeding	NOUN
fcis-9479	118	19	of	of	ADP
fcis-9479	118	20	the	the	DET
fcis-9479	118	21	2012	2012	NUM
fcis-9479	118	22	siam	siam	ADJ
fcis-9479	118	23	international	international	ADJ
fcis-9479	118	24	conference	conference	NOUN
fcis-9479	118	25	on	on	ADP
fcis-9479	118	26	data	datum	NOUN
fcis-9479	118	27	mining	mining	NOUN
fcis-9479	118	28	.	.	PUNCT
fcis-9479	119	1	siam	siam	PROPN
fcis-9479	119	2	,	,	PUNCT
fcis-9479	119	3	2012	2012	NUM
fcis-9479	119	4	:	:	PUNCT
fcis-9479	119	5	153	153	NUM
fcis-9479	119	6	-	-	SYM
fcis-9479	119	7	164	164	NUM
fcis-9479	119	8	.	.	PUNCT
fcis-9479	120	1	[	[	X
fcis-9479	120	2	8	8	NUM
fcis-9479	120	3	]	]	X
fcis-9479	120	4	zhao	zhao	PROPN
fcis-9479	120	5	l	l	PROPN
fcis-9479	120	6	,	,	PUNCT
fcis-9479	120	7	hu	hu	PROPN
fcis-9479	120	8	q	q	PROPN
fcis-9479	120	9	,	,	PUNCT
fcis-9479	120	10	wang	wang	PROPN
fcis-9479	120	11	w.	w.	PROPN
fcis-9479	120	12	heterogeneous	heterogeneous	ADJ
fcis-9479	120	13	feature	feature	NOUN
fcis-9479	120	14	selection	selection	NOUN
fcis-9479	120	15	with	with	ADP
fcis-9479	120	16	multi	multi	ADJ
fcis-9479	120	17	-	-	ADJ
fcis-9479	120	18	modal	modal	ADJ
fcis-9479	120	19	deep	deep	ADJ
fcis-9479	120	20	neural	neural	ADJ
fcis-9479	120	21	networks	network	NOUN
fcis-9479	120	22	and	and	CCONJ
fcis-9479	120	23	sparse	sparse	ADJ
fcis-9479	120	24	group	group	NOUN
fcis-9479	120	25	lasso	lasso	NOUN
fcis-9479	121	1	[	[	X
fcis-9479	121	2	j	j	X
fcis-9479	121	3	]	]	X
fcis-9479	121	4	.	.	PUNCT
fcis-9479	122	1	ieee	ieee	NOUN
fcis-9479	122	2	transactions	transaction	NOUN
fcis-9479	122	3	on	on	ADP
fcis-9479	122	4	multimedia	multimedia	NOUN
fcis-9479	122	5	,	,	PUNCT
fcis-9479	122	6	2015	2015	NUM
fcis-9479	122	7	,	,	PUNCT
fcis-9479	122	8	17(11	17(11	NUM
fcis-9479	122	9	):	):	PUNCT
fcis-9479	122	10	1936	1936	NUM
fcis-9479	122	11	-	-	SYM
fcis-9479	122	12	19	19	NUM
fcis-9479	122	13	48	48	NUM
fcis-9479	122	14	.	.	PUNCT
fcis-9479	123	1	[	[	X
fcis-9479	123	2	9	9	NUM
fcis-9479	123	3	]	]	X
fcis-9479	123	4	joulin	joulin	NOUN
fcis-9479	123	5	a	a	PRON
fcis-9479	123	6	,	,	PUNCT
fcis-9479	123	7	grave	grave	ADJ
fcis-9479	123	8	e	e	NOUN
fcis-9479	123	9	,	,	PUNCT
fcis-9479	123	10	bojanowski	bojanowski	NOUN
fcis-9479	123	11	p	p	X
fcis-9479	123	12	,	,	PUNCT
fcis-9479	123	13	mikolov	mikolov	ADJ
fcis-9479	123	14	t.	t.	NOUN
fcis-9479	123	15	bag	bag	NOUN
fcis-9479	123	16	of	of	ADP
fcis-9479	123	17	tricks	trick	NOUN
fcis-9479	123	18	for	for	ADP
fcis-9479	123	19	efficient	efficient	ADJ
fcis-9479	123	20	text	text	NOUN
fcis-9479	123	21	classification[j	classification[j	NOUN
fcis-9479	123	22	]	]	PUNCT
fcis-9479	123	23	.	.	PUNCT
fcis-9479	124	1	arxiv	arxiv	PROPN
fcis-9479	124	2	preprint	preprint	VERB
fcis-9479	124	3	arxiv:1607.01759	arxiv:1607.01759	NOUN
fcis-9479	124	4	,	,	PUNCT
fcis-9479	124	5	2016	2016	NUM
fcis-9479	124	6	.	.	PUNCT
fcis-9479	125	1	[	[	X
fcis-9479	125	2	10	10	NUM
fcis-9479	125	3	]	]	X
fcis-9479	125	4	wang	wang	PROPN
fcis-9479	125	5	y	y	PROPN
fcis-9479	125	6	,	,	PUNCT
fcis-9479	125	7	ma	ma	PROPN
fcis-9479	125	8	f	f	PROPN
fcis-9479	125	9	,	,	PUNCT
fcis-9479	125	10	jin	jin	PROPN
fcis-9479	125	11	z	z	PROPN
fcis-9479	125	12	,	,	PUNCT
fcis-9479	125	13	yuan	yuan	PROPN
fcis-9479	125	14	y	y	PROPN
fcis-9479	125	15	,	,	PUNCT
fcis-9479	125	16	xun	xun	PROPN
fcis-9479	125	17	g	g	PROPN
fcis-9479	125	18	,	,	PUNCT
fcis-9479	125	19	jha	jha	PROPN
fcis-9479	125	20	k	k	PROPN
fcis-9479	125	21	,	,	PUNCT
fcis-9479	125	22	su	su	PROPN
fcis-9479	125	23	l	l	PROPN
fcis-9479	125	24	,	,	PUNCT
fcis-9479	125	25	gao	gao	PROPN
fcis-9479	125	26	j.	j.	PROPN
fcis-9479	125	27	eann	eann	PROPN
fcis-9479	125	28	:	:	PUNCT
fcis-9479	125	29	event	event	NOUN
fcis-9479	125	30	adversarial	adversarial	ADJ
fcis-9479	125	31	neural	neural	ADJ
fcis-9479	125	32	networks	network	NOUN
fcis-9479	125	33	for	for	ADP
fcis-9479	125	34	multimodal	multimodal	ADJ
fcis-9479	125	35	fake	fake	ADJ
fcis-9479	125	36	news	news	NOUN
fcis-9479	125	37	detection[c]//proceedings	detection[c]//proceeding	NOUN
fcis-9479	125	38	of	of	ADP
fcis-9479	125	39	the	the	DET
fcis-9479	125	40	24th	24th	ADJ
fcis-9479	125	41	acm	acm	PROPN
fcis-9479	125	42	sigkdd	sigkdd	NOUN
fcis-9479	125	43	international	international	ADJ
fcis-9479	125	44	conference	conference	NOUN
fcis-9479	125	45	on	on	ADP
fcis-9479	125	46	knowledge	knowledge	PROPN
fcis-9479	125	47	discovery	discovery	PROPN
fcis-9479	125	48	&	&	CCONJ
fcis-9479	125	49	data	data	PROPN
fcis-9479	125	50	mining	mining	NOUN
fcis-9479	125	51	.	.	PUNCT
fcis-9479	126	1	2018	2018	NUM
fcis-9479	126	2	:	:	PUNCT
fcis-9479	126	3	849	849	NUM
fcis-9479	126	4	-	-	SYM
fcis-9479	126	5	857	857	NUM
fcis-9479	126	6	.	.	PUNCT
fcis-9479	127	1	[	[	X
fcis-9479	127	2	11	11	NUM
fcis-9479	127	3	]	]	PUNCT
fcis-9479	127	4	khattar	khattar	NOUN
fcis-9479	127	5	d	d	PROPN
fcis-9479	127	6	,	,	PUNCT
fcis-9479	127	7	goud	goud	PROPN
fcis-9479	127	8	j	j	PROPN
fcis-9479	127	9	s	s	PROPN
fcis-9479	127	10	,	,	PUNCT
fcis-9479	127	11	gupta	gupta	PROPN
fcis-9479	127	12	m	m	PROPN
fcis-9479	127	13	,	,	PUNCT
fcis-9479	127	14	varma	varma	PROPN
fcis-9479	127	15	v.	v.	ADP
fcis-9479	127	16	mvae	mvae	PROPN
fcis-9479	127	17	:	:	PUNCT
fcis-9479	127	18	multimodal	multimodal	ADJ
fcis-9479	127	19	variational	variational	ADJ
fcis-9479	127	20	autoencoder	autoencoder	NOUN
fcis-9479	127	21	for	for	ADP
fcis-9479	127	22	fake	fake	ADJ
fcis-9479	127	23	news	news	NOUN
fcis-9479	127	24	detection	detection	NOUN
fcis-9479	127	25	[	[	X
fcis-9479	127	26	c	c	X
fcis-9479	127	27	]	]	X
fcis-9479	127	28	//	//	NOUN
fcis-9479	127	29	the	the	DET
fcis-9479	127	30	world	world	NOUN
fcis-9479	127	31	wide	wide	ADJ
fcis-9479	127	32	web	web	NOUN
fcis-9479	127	33	conference	conference	NOUN
fcis-9479	127	34	.	.	PUNCT
fcis-9479	128	1	2019	2019	NUM
fcis-9479	128	2	:	:	PUNCT
fcis-9479	128	3	2915	2915	NUM
fcis-9479	128	4	-	-	SYM
fcis-9479	128	5	2921	2921	NUM
fcis-9479	128	6	.	.	PUNCT
fcis-9479	129	1	[	[	X
fcis-9479	129	2	12	12	NUM
fcis-9479	129	3	]	]	PUNCT
fcis-9479	129	4	church	church	NOUN
fcis-9479	129	5	k	k	PROPN
fcis-9479	129	6	,	,	PUNCT
fcis-9479	129	7	hanks	hanks	PROPN
fcis-9479	129	8	p.	p.	PROPN
fcis-9479	129	9	word	word	PROPN
fcis-9479	129	10	association	association	NOUN
fcis-9479	129	11	norms	norm	NOUN
fcis-9479	129	12	,	,	PUNCT
fcis-9479	129	13	mutual	mutual	ADJ
fcis-9479	129	14	information	information	NOUN
fcis-9479	129	15	,	,	PUNCT
fcis-9479	129	16	and	and	CCONJ
fcis-9479	129	17	lexicography[j	lexicography[j	PROPN
fcis-9479	129	18	]	]	PUNCT
fcis-9479	129	19	.	.	PUNCT
fcis-9479	130	1	computational	computational	ADJ
fcis-9479	130	2	linguistics	linguistic	NOUN
fcis-9479	130	3	,	,	PUNCT
fcis-9479	130	4	1990	1990	NUM
fcis-9479	130	5	,	,	PUNCT
fcis-9479	130	6	16(1	16(1	NUM
fcis-9479	130	7	):	):	PUNCT
fcis-9479	130	8	22	22	NUM
fcis-9479	130	9	-	-	SYM
fcis-9479	130	10	29	29	NUM
fcis-9479	130	11	.	.	PUNCT
fcis-9479	131	1	[	[	X
fcis-9479	131	2	13	13	NUM
fcis-9479	131	3	]	]	X
fcis-9479	131	4	singhal	singhal	PROPN
fcis-9479	131	5	s	s	PROPN
fcis-9479	131	6	,	,	PUNCT
fcis-9479	131	7	shah	shah	NOUN
fcis-9479	131	8	r	r	NOUN
fcis-9479	131	9	r	r	PROPN
fcis-9479	131	10	,	,	PUNCT
fcis-9479	131	11	chakraborty	chakraborty	PROPN
fcis-9479	131	12	t	t	PROPN
fcis-9479	131	13	,	,	PUNCT
fcis-9479	131	14	kumaraguru	kumaraguru	VERB
fcis-9479	131	15	p	p	NOUN
fcis-9479	131	16	,	,	PUNCT
fcis-9479	131	17	satoh	satoh	NOUN
fcis-9479	131	18	s.	s.	PROPN
fcis-9479	131	19	spotfake	spotfake	PROPN
fcis-9479	131	20	:	:	PUNCT
fcis-9479	131	21	a	a	DET
fcis-9479	131	22	multi	multi	ADJ
fcis-9479	131	23	-	-	ADJ
fcis-9479	131	24	modal	modal	ADJ
fcis-9479	131	25	framework	framework	NOUN
fcis-9479	131	26	for	for	ADP
fcis-9479	131	27	fake	fake	ADJ
fcis-9479	131	28	news	news	NOUN
fcis-9479	131	29	detection[c]//2019	detection[c]//2019	PUNCT
fcis-9479	131	30	ieee	ieee	NOUN
fcis-9479	131	31	fifth	fifth	ADJ
fcis-9479	131	32	international	international	ADJ
fcis-9479	131	33	conference	conference	NOUN
fcis-9479	131	34	on	on	ADP
fcis-9479	131	35	multimedia	multimedia	PROPN
fcis-9479	131	36	big	big	ADJ
fcis-9479	131	37	data	datum	NOUN
fcis-9479	131	38	(	(	PUNCT
fcis-9479	131	39	bigmm	bigmm	NOUN
fcis-9479	131	40	)	)	PUNCT
fcis-9479	131	41	.	.	PUNCT
fcis-9479	132	1	ieee	ieee	NOUN
fcis-9479	132	2	,	,	PUNCT
fcis-9479	132	3	2019	2019	NUM
fcis-9479	132	4	:	:	PUNCT
fcis-9479	132	5	39	39	NUM
fcis-9479	132	6	-	-	SYM
fcis-9479	132	7	47	47	NUM
fcis-9479	132	8	.	.	PUNCT
fcis-9479	133	1	[	[	X
fcis-9479	133	2	14	14	NUM
fcis-9479	133	3	]	]	X
fcis-9479	133	4	devlin	devlin	PROPN
fcis-9479	133	5	j	j	PROPN
fcis-9479	133	6	,	,	PUNCT
fcis-9479	133	7	chang	chang	PROPN
fcis-9479	133	8	m	m	PROPN
fcis-9479	133	9	w	w	PROPN
fcis-9479	133	10	,	,	PUNCT
fcis-9479	133	11	lee	lee	PROPN
fcis-9479	133	12	k	k	PROPN
fcis-9479	133	13	,	,	PUNCT
fcis-9479	133	14	toutanova	toutanova	PROPN
fcis-9479	133	15	k.	k.	PROPN
fcis-9479	133	16	bert	bert	PROPN
fcis-9479	133	17	:	:	PUNCT
fcis-9479	133	18	pre	pre	ADJ
fcis-9479	133	19	-	-	NOUN
fcis-9479	133	20	training	training	NOUN
fcis-9479	133	21	of	of	ADP
fcis-9479	133	22	deep	deep	ADJ
fcis-9479	133	23	bidirectional	bidirectional	ADJ
fcis-9479	133	24	transformers	transformer	NOUN
fcis-9479	133	25	for	for	ADP
fcis-9479	133	26	language	language	NOUN
fcis-9479	133	27	understanding	understanding	NOUN
fcis-9479	133	28	[	[	X
fcis-9479	133	29	j	j	X
fcis-9479	133	30	]	]	X
fcis-9479	133	31	.	.	PUNCT
fcis-9479	134	1	arxiv	arxiv	PROPN
fcis-9479	134	2	preprint	preprint	PROPN
fcis-9479	134	3	arxiv:1810.04805	arxiv:1810.04805	PROPN
fcis-9479	134	4	,	,	PUNCT
fcis-9479	134	5	2018	2018	NUM
fcis-9479	134	6	.	.	PUNCT
fcis-9479	135	1	[	[	X
fcis-9479	135	2	15	15	NUM
fcis-9479	135	3	]	]	X
fcis-9479	135	4	deng	deng	PROPN
fcis-9479	135	5	j	j	PROPN
fcis-9479	135	6	,	,	PUNCT
fcis-9479	135	7	dong	dong	PROPN
fcis-9479	135	8	w	w	PROPN
fcis-9479	135	9	,	,	PUNCT
fcis-9479	135	10	socher	socher	NOUN
fcis-9479	135	11	r	r	NOUN
fcis-9479	135	12	,	,	PUNCT
fcis-9479	135	13	li	li	PROPN
fcis-9479	135	14	l	l	PROPN
fcis-9479	135	15	j	j	PROPN
fcis-9479	135	16	,	,	PUNCT
fcis-9479	135	17	li	li	PROPN
fcis-9479	135	18	k	k	PROPN
fcis-9479	135	19	,	,	PUNCT
fcis-9479	135	20	fei	fei	PROPN
fcis-9479	135	21	-	-	PUNCT
fcis-9479	135	22	fei	fei	PROPN
fcis-9479	135	23	l.	l.	PROPN
fcis-9479	135	24	imagenet	imagenet	PROPN
fcis-9479	135	25	:	:	PUNCT
fcis-9479	135	26	a	a	DET
fcis-9479	135	27	large	large	ADJ
fcis-9479	135	28	-	-	PUNCT
fcis-9479	135	29	scale	scale	NOUN
fcis-9479	135	30	hierarchical	hierarchical	ADJ
fcis-9479	135	31	image	image	NOUN
fcis-9479	135	32	database[c]//2009	database[c]//2009	PROPN
fcis-9479	135	33	ieee	ieee	NOUN
fcis-9479	135	34	conference	conference	NOUN
fcis-9479	135	35	on	on	ADP
fcis-9479	135	36	computer	computer	NOUN
fcis-9479	135	37	vision	vision	NOUN
fcis-9479	135	38	and	and	CCONJ
fcis-9479	135	39	pattern	pattern	NOUN
fcis-9479	135	40	recognition	recognition	NOUN
fcis-9479	135	41	.	.	PUNCT
fcis-9479	136	1	ieee	ieee	PROPN
fcis-9479	136	2	,	,	PUNCT
fcis-9479	136	3	2009	2009	NUM
fcis-9479	136	4	:	:	PUNCT
fcis-9479	136	5	248	248	NUM
fcis-9479	136	6	-	-	SYM
fcis-9479	136	7	255	255	NUM
fcis-9479	136	8	.	.	PUNCT
fcis-9479	137	1	[	[	X
fcis-9479	137	2	16	16	NUM
fcis-9479	137	3	]	]	X
fcis-9479	137	4	singhal	singhal	PROPN
fcis-9479	137	5	s	s	PROPN
fcis-9479	137	6	,	,	PUNCT
fcis-9479	137	7	kabra	kabra	PROPN
fcis-9479	137	8	a	a	PRON
fcis-9479	137	9	,	,	PUNCT
fcis-9479	137	10	sharma	sharma	PROPN
fcis-9479	137	11	m	m	PROPN
fcis-9479	137	12	,	,	PUNCT
fcis-9479	137	13	shah	shah	PROPN
fcis-9479	137	14	r	r	PROPN
fcis-9479	137	15	r	r	PROPN
fcis-9479	137	16	,	,	PUNCT
fcis-9479	137	17	chakraborty	chakraborty	PROPN
fcis-9479	137	18	t	t	PROPN
fcis-9479	137	19	,	,	PUNCT
fcis-9479	137	20	kumaraguru	kumaraguru	PROPN
fcis-9479	137	21	p.	p.	PROPN
fcis-9479	138	1	spotfake+	spotfake+	PROPN
fcis-9479	138	2	:	:	PUNCT
fcis-9479	138	3	a	a	DET
fcis-9479	138	4	multimodal	multimodal	ADJ
fcis-9479	138	5	framework	framework	NOUN
fcis-9479	138	6	for	for	ADP
fcis-9479	138	7	fake	fake	ADJ
fcis-9479	138	8	news	news	NOUN
fcis-9479	138	9	detection	detection	NOUN
fcis-9479	138	10	via	via	ADP
fcis-9479	138	11	transfer	transfer	NOUN
fcis-9479	138	12	learning	learning	NOUN
fcis-9479	138	13	(	(	PUNCT
fcis-9479	138	14	student	student	NOUN
fcis-9479	138	15	abstract	abstract	NOUN
fcis-9479	138	16	)	)	PUNCT
fcis-9479	139	1	[	[	X
fcis-9479	139	2	c]//proceedings	c]//proceeding	NOUN
fcis-9479	139	3	of	of	ADP
fcis-9479	139	4	the	the	DET
fcis-9479	139	5	aaai	aaai	PROPN
fcis-9479	139	6	conference	conference	NOUN
fcis-9479	139	7	on	on	ADP
fcis-9479	139	8	artificial	artificial	ADJ
fcis-9479	139	9	intelligence	intelligence	NOUN
fcis-9479	139	10	:	:	PUNCT
fcis-9479	139	11	34	34	NUM
fcis-9479	139	12	.	.	PUNCT
fcis-9479	139	13	2020	2020	NUM
fcis-9479	139	14	:	:	PUNCT
fcis-9479	139	15	13915	13915	NUM
fcis-9479	139	16	-	-	SYM
fcis-9479	139	17	13916	13916	NUM
fcis-9479	139	18	.	.	PUNCT
fcis-9479	140	1	[	[	X
fcis-9479	140	2	17	17	NUM
fcis-9479	140	3	]	]	X
fcis-9479	140	4	yang	yang	PROPN
fcis-9479	140	5	z	z	PROPN
fcis-9479	140	6	,	,	PUNCT
fcis-9479	140	7	dai	dai	PROPN
fcis-9479	140	8	z	z	PROPN
fcis-9479	140	9	,	,	PUNCT
fcis-9479	140	10	yang	yang	PROPN
fcis-9479	140	11	y	y	PROPN
fcis-9479	140	12	,	,	PUNCT
fcis-9479	140	13	carbonell	carbonell	PROPN
fcis-9479	140	14	j	j	PROPN
fcis-9479	140	15	,	,	PUNCT
fcis-9479	140	16	salakhutdinov	salakhutdinov	NOUN
fcis-9479	140	17	r	r	NOUN
fcis-9479	140	18	r	r	PROPN
fcis-9479	140	19	,	,	PUNCT
fcis-9479	140	20	le	le	X
fcis-9479	140	21	q	q	PROPN
fcis-9479	140	22	v.	v.	ADP
fcis-9479	140	23	xlnet	xlnet	PROPN
fcis-9479	140	24	:	:	PUNCT
fcis-9479	140	25	generalized	generalize	VERB
fcis-9479	140	26	autoregressive	autoregressive	ADJ
fcis-9479	140	27	pretraining	pretraine	VERB
fcis-9479	140	28	for	for	ADP
fcis-9479	140	29	language	language	NOUN
fcis-9479	140	30	understanding[j	understanding[j	NOUN
fcis-9479	140	31	]	]	PUNCT
fcis-9479	140	32	.	.	PUNCT
fcis-9479	141	1	advances	advance	NOUN
fcis-9479	141	2	in	in	ADP
fcis-9479	141	3	neural	neural	ADJ
fcis-9479	141	4	information	information	NOUN
fcis-9479	141	5	processing	processing	NOUN
fcis-9479	141	6	systems	system	NOUN
fcis-9479	141	7	,	,	PUNCT
fcis-9479	141	8	2019	2019	NUM
fcis-9479	141	9	,	,	PUNCT
fcis-9479	141	10	32	32	NUM
fcis-9479	141	11	.	.	PUNCT
fcis-9479	142	1	[	[	X
fcis-9479	142	2	18	18	NUM
fcis-9479	142	3	]	]	X
fcis-9479	142	4	zhou	zhou	PROPN
fcis-9479	142	5	x	x	PROPN
fcis-9479	142	6	,	,	PUNCT
fcis-9479	142	7	wu	wu	PROPN
fcis-9479	142	8	j	j	PROPN
fcis-9479	142	9	,	,	PUNCT
fcis-9479	142	10	zafarani	zafarani	PROPN
fcis-9479	142	11	r.	r.	PROPN
fcis-9479	142	12	safe	safe	ADJ
fcis-9479	142	13	:	:	PUNCT
fcis-9479	142	14	similarity	similarity	NOUN
fcis-9479	142	15	-	-	PUNCT
fcis-9479	142	16	aware	aware	ADJ
fcis-9479	142	17	multimodal	multimodal	ADJ
fcis-9479	142	18	fake	fake	ADJ
fcis-9479	142	19	news	news	NOUN
fcis-9479	142	20	detection	detection	NOUN
fcis-9479	142	21	(	(	PUNCT
fcis-9479	142	22	2020	2020	NUM
fcis-9479	142	23	)	)	PUNCT
fcis-9479	143	1	[	[	X
fcis-9479	143	2	j	j	X
fcis-9479	143	3	]	]	X
fcis-9479	143	4	.	.	PUNCT
fcis-9479	144	1	preprint	preprint	NOUN
fcis-9479	144	2	.	.	PUNCT
fcis-9479	145	1	arxiv	arxiv	PROPN
fcis-9479	145	2	,	,	PUNCT
fcis-9479	145	3	2020	2020	NUM
fcis-9479	145	4	,	,	PUNCT
fcis-9479	145	5	200304981	200304981	NUM
fcis-9479	145	6	:	:	PUNCT
fcis-9479	146	1	2	2	X
fcis-9479	146	2	.	.	PUNCT
fcis-9479	147	1	[	[	X
fcis-9479	147	2	19	19	NUM
fcis-9479	147	3	]	]	PUNCT
fcis-9479	147	4	hu	hu	PROPN
fcis-9479	147	5	g	g	PROPN
fcis-9479	147	6	,	,	PUNCT
fcis-9479	147	7	ding	de	VERB
fcis-9479	147	8	y	y	PROPN
fcis-9479	147	9	,	,	PUNCT
fcis-9479	147	10	qi	qi	PROPN
fcis-9479	147	11	s	s	PROPN
fcis-9479	147	12	,	,	PUNCT
fcis-9479	147	13	wang	wang	PROPN
fcis-9479	147	14	x	x	PROPN
fcis-9479	147	15	,	,	PUNCT
fcis-9479	147	16	liao	liao	PROPN
fcis-9479	147	17	q.	q.	PROPN
fcis-9479	147	18	multi	multi	ADJ
fcis-9479	147	19	-	-	ADJ
fcis-9479	147	20	depth	depth	ADJ
fcis-9479	147	21	graph	graph	NOUN
fcis-9479	147	22	convolutional	convolutional	ADJ
fcis-9479	147	23	networks	network	NOUN
fcis-9479	147	24	for	for	ADP
fcis-9479	147	25	fake	fake	ADJ
fcis-9479	147	26	news	news	NOUN
fcis-9479	147	27	detection[c]//natural	detection[c]//natural	PROPN
fcis-9479	147	28	language	language	NOUN
fcis-9479	147	29	processing	processing	NOUN
fcis-9479	147	30	and	and	CCONJ
fcis-9479	147	31	chinese	chinese	ADJ
fcis-9479	147	32	computing	computing	NOUN
fcis-9479	147	33	:	:	PUNCT
fcis-9479	147	34	8th	8th	ADJ
fcis-9479	147	35	ccf	ccf	ADP
fcis-9479	147	36	international	international	ADJ
fcis-9479	147	37	conference	conference	PROPN
fcis-9479	147	38	,	,	PUNCT
fcis-9479	147	39	nlpcc	nlpcc	NOUN
fcis-9479	147	40	2019	2019	NUM
fcis-9479	147	41	,	,	PUNCT
fcis-9479	147	42	dunhuang	dunhuang	PROPN
fcis-9479	147	43	,	,	PUNCT
fcis-9479	147	44	china	china	PROPN
fcis-9479	147	45	,	,	PUNCT
fcis-9479	147	46	october	october	PROPN
fcis-9479	147	47	9–14	9–14	PROPN
fcis-9479	147	48	,	,	PUNCT
fcis-9479	147	49	2019	2019	NUM
fcis-9479	147	50	,	,	PUNCT
fcis-9479	147	51	proceedings	proceeding	NOUN
fcis-9479	147	52	,	,	PUNCT
fcis-9479	147	53	part	part	NOUN
fcis-9479	147	54	i	i	PRON
fcis-9479	147	55	8	8	NUM
fcis-9479	147	56	.	.	PUNCT
fcis-9479	147	57	springer	springer	NOUN
fcis-9479	147	58	,	,	PUNCT
fcis-9479	147	59	2019	2019	NUM
fcis-9479	147	60	:	:	PUNCT
fcis-9479	147	61	698	698	NUM
fcis-9479	147	62	-	-	SYM
fcis-9479	147	63	710	710	NUM
fcis-9479	147	64	.	.	PUNCT