Table 9 Performance comparison of different models on a clinical dataset of impacted wisdom teeth
From: An imConvNet-based deep learning model for Chinese medical named entity recognition
Models | P (precision) | R (recall) | F1-score |
|---|---|---|---|
IDCNN-CRF | 82.97 | 89.62 | 86.17 |
BiLSTM-CRF | 84.22 | 88.41 | 86.26 |
imConvNet-CRF | 83.01 | 92.22 | 87.37 |
imConvNet-BiLSTM-CRF | 86.80 | 91.51 | 89.09 |
BERT-imConvNet-CRF | 89.80 | 95.75 | 92.68 |
BERT-imConvNet-BiLSTM-CRF | 91.61 | 96.30 | 93.89 |