Named entity recognition for Chinese judgment documents based on BiLSTM and CRF

EURASIP Journal on Image and Video Processing

Table 6 Our method and other methods are tested in our dataset and the results are compared

Model	Precision	Recall	F1
Multiple HMMs [24]	71.64	71.06	71.34
Char baseline	68.79	60.35	64.30
+bichar+softword LSTM	74.36	69.43	71.81
Lattice LSTM [5]	76.35	71.56	73.88
Segment-level neural network [25]	71.60	69.40	70.48
BiLSTM-CRF	71.14	72.21	71.67
Our method	77.08	73.69	75.35