A Multitask Deep Learning Framework for DNER

Table 8

Performance comparison of adopting multitask learning.

		Precision	Recall	F1
DDI2011	Single-task	91.13	89.51	90.30
	Multitask	94.36	92.13	93.22
	Δ	+3.23	+2.62	+2.92
DDI2013	Single-task	83.42	78.01	80.62
	Multitask	85.56	81.45	83.45
	Δ	+2.14	+3.44	+2.83