Skip to main content

Table 9 Results of different NER models on CCKS-2019

From: Named entity recognition of Chinese electronic medical records based on a hybrid neural network and medical MC-BERT

Models

Precision

Recall

F1

Cost time

BiLSTM + CRF

81.11

80.47

80.79

6 h 23 m

RD + CNN + CRF [31]

81.87

82.03

81.95

5 h 43 m

ELMO-BiLSTM-CRF [32]

82.31

81.89

81.10

5 h 33 m

BERT-BiLSTM-CRF

82.09

87.32

84.62

4 h 26 m

BERT-wwm + BiLSTM + CRF [33]

82.47

87.42

84.87

4 h47 m

MC_BERT-BiLSTM-CRF

83.10

87.42

85.20

4 h 13 m

MC_BERT-BiLSTM-MHA-CRF

83.24

87.93

85.52

4 h 39 m

MC_BERT-BiLSTM-CNN-CRF

83.04

87.77

85.34

4 h 32 m

MC_BERT-BiLSTM-CNN-MHA-CRF

84.90

87.67

86.27

4 h 48 m