GPDminer: a tool for extracting named entities and analyzing relations in biological literature

BMC Bioinformatics

Table 5 Comparison of NER models for genes/proteins, diseases

Entity types	Models	Precision	Recall	F1-score
	Ours	0.8933	0.9300	0.9112
	TaggerOne Joint [40]	0.8510	0.8080	0.8290
	TaggerOne NER-only [40]	0.8350	0.7960	0.8150
Disease	DNorm [36]	0.8030	0.7630	0.8720
	Sachan et al. [41]	0.8641	0.8831	0.8734
	CollaboNet [42]	0.8548	0.8727	0.8636
	LSTM-CRF (iii) of Habibi et al. [43]	0.8531	0.8358	0.8444
	Ours	0.8551	0.8899	0.8719
	GNormPlus [37]	0.7840	0.7920	0.7880
Gene/protein	Sachan et al. [41]	0.8181	0.8157	0.8169
	CollaboNet [42]	0.8049	0.7899	0.7973
	LSTM-CRF (iii) of Habibi et al. [43]	0.7750	0.7813	0.7782

ISSN: 1471-2105