Table 5.

The microevaluation results on the development and test sets under the optimal mode for the 3 machine learning algorithms

Dataset	Algorithm	P	R	F
Development	CRF	0.937	0.892	0.914
	C-BiLSTM-CRF	0.929	0.913	0.921
	L-BiLSTM-CRF	0.934	0.907	0.920
Test set	CRF	0.946	0.871	0.907
	C-BiLSTM-CRF	0.935	0.892	0.913
	L-BiLSTM-CRF	0.937	0.890	0.913

C-BiLSTM-CRF: convolutional neural network bidirectional long short-term memory conditional random field; CRF: conditional random field; F: micro F-score; L-BiLSTM-CRF: long short-term memory bidirectional long short-term memory conditional random field P: precision; R: recall.