Multilayer networks for text analysis with multiple data types

Hyland, Charles C.; Tao, Yuanming; Azizi, Lamiae; Gerlach, Martin; Peixoto, Tiago P.; Altmann, Eduardo G.

doi:10.1140/epjds/s13688-021-00288-5

EPJ Data Science

Table 3 Groups of word types as topics. Upper table: the 20 most frequent words in the 12 topics (word groups) found in the consensus partition of the Hyperlink + Text model. Lower table: the topic proportion (8) of the four groups of documents

From: Multilayer networks for text analysis with multiple data types

Group	Topic 1	Topic 2	Topic 3	Topic 4	Topic 5	Topic 6	Topic 7	Topic 8	Topic 9	Topic 10	Topic 11	Topic 12
	occur	chemical	protein	one	system	group	transform	action	happen	quantum	every	factor
	specific	loop	cell	also	time	space	law	translation	frequency	equation	basis	expression
	within	electron	site	use	state	define	critical	class	fast	transformation	unique	region
	process	transition	activity	form	work	field	magnetic	line	output	theorem	degree	molecular
	produce	chemistry	pathway	function	second	point	matter	measure	input	symmetry	positive	sequence
	increase	atom	cellular	give	effect	theory	scale	central	algorithm	dimension	open	information
	mechanism	stable	organism	two	phase	value	plane	primary	digital	denote	close	end
	cause	ion	amino	may	interaction	product	spectrum	unknown		classical	fix	molecule
	acid	pattern	enzyme	however	potential	constant	volume	interior		let	reference	domain
	target	crystal	synthesis	example	could	vector	temperature	surface		mechanic	exact	signal
	step	reach	species	result	derive	physic	axis	side		coordinate	closed	alternative
	bind	prediction	tissue	call	difference	physical	statistical	center		lie	restrict	bond
	gene	label	proteins	first	full	energy	sign	flow		real	picture	family
	control	compound	cancer	different	free	element	effective	copy		mathematics	infinitesimal	interact
	human	reaction	activate	know	hold	particle	symbol	block		geometry	index	cycle
	encode	crystallography	genetic	make	observe	parameter	speed	configuration		operator	interpretation	biological
	body	oxygen	membrane	number	research	whose	electric	read		differential	formulate	life
	dna	solid	release	structure	propose	mathematical	equilibrium	alternate		representation	equivalently	translate
	rate	unstable	mutation	show	year	linear	unchanged	table		invariant	maximal	structural
	formation	atomic	regulation	include	paper	matrix	assumption	transport		generalize	scheme	biology
1 (Mathematics)	0.1481	0.0213	0.0444	0.400	0.0578	0.0722	0.0510	0.0127	0.00163	0.147	0.0105	0.0338
2 (Physics)	0.138	0.0409	0.0291	0.404	0.0643	0.0710	0.07940	0.0127	0.0009583	0.105	0.0102	0.0445
3 (Biology)	0.297	0.0234	0.0536	0.351	0.0673	0.0426	0.0313	0.0121	0.00166	0.0347	0.00929	0.0759
4 (New Group)	0.107	0.0342	0.0355	0.405	0.0771	0.0728	0.0895	0.0122	0.00156	0.112	0.0113	0.0428

Back to article page