Ensemble of deep learning language models to support the creation of living systematic reviews for the COVID-19 literature

Table 1 Dataset document count and proportion by class, subclass, and sub-subclass

Class	Subclass	Sub-subclass	Count	%
ORIGINAL	EPI	Case report	241	3.8
ORIGINAL	EPI	Case series	350	5.5
ORIGINAL	EPI	Case–control study	74	1.2
ORIGINAL	EPI	Cohort study	246	3.9
ORIGINAL	EPI	Cross-sectional study	284	4.5
ORIGINAL	EPI	Diagnostic study	181	2.8
ORIGINAL	EPI	Ecological study	92	1.4
ORIGINAL	EPI	Guidelines	326	5.1
ORIGINAL	EPI	Modelling study	808	12.7
ORIGINAL	EPI	Other	130	2.0
ORIGINAL	EPI	Outbreak or surveillance report	133	2.1
ORIGINAL	EPI	Qualitative study	35	0.5
ORIGINAL	EPI	Review	725	11.4
ORIGINAL	EPI	Trial	40	0.6
ORIGINAL	BASIC	Animal experiment	43	0.7
ORIGINAL	BASIC	Basic research review	135	2.1
ORIGINAL	BASIC	Biochemical/protein structure studies	264	4.1
ORIGINAL	BASIC	In vitro experiment	85	1.3
ORIGINAL	BASIC	Sequencing and phylogenetics	241	3.8
ORIGINAL	BASIC	Within-host modeling	31	0.5
NON-ORIGINAL	OTHER	Other	143	2.2
NON-ORIGINAL	OTHER	Comment, editorial, …, non-original	1758	27.6
		ALL	6365	100.0

ISSN: 2046-4053