A neural network multi-task learning approach to biomedical named entity recognition

BMC Bioinformatics

Table 9 Single task and multi-task f-scores on NER tasks

Dataset	Baseline	STM	MO-MTM	D-MTM
AnatEM	81.79	81.55	81.83	82.21*
BC2GM	70.31	72.63	73.17	72.87
BC4CHEMD	81.08	82.95	82.37	83.02
BC5CDR	83.11	83.66	83.90	83.83
BioNLP09	81.84	83.90	84.20	84.10
BioNLP11EPI	74.98	77.72	78.86*	78.03*
BioNLP11ID	81.44	81.50	80.58*	81.73
BioNLP13CG	75.23	76.74	78.90*	77.52*
BioNLP13GE	72.49	73.28	78.58*	74.00*
BioNLP13PC	79.35	80.61	81.92*	81.50*
CRAFT	78.76	79.55	79.10	79.56
Ex-PTM	65.75	68.56	74.90*	69.67*
JNLPBA	67.43	69.60	70.09	69.44
Linnaeus	79.01	83.98	81.57	84.04
NCBI-Disease	79.09	80.26	79.02	80.37
Average	76.78	78.43	79.26	78.79

(Bold: best scores, *: statistically significant compared to single-task model)

ISSN: 1471-2105