Table 7 Experimental results at ERTCMM. where (baseline) is the model comparison baseline, (LSTM) means that the method is based on LSTM
From: Chinese medical entity recognition based on the dual-branch TENER model
Model | P(%) | R(%) | F1 score(%) |
|---|---|---|---|
RoBERT-CRF(baseline) | 65.79 | 73.97 | 69.64 |
RoBERT-BiLSTM- | Â | Â | Â |
CRF(baseline) | 67.56 | 73.36 | 70.34 |
Lattice LSTM | 60.52 | 76.15 | 67.44 |
Soft-Lexicon(LSTM) | 62.27 | 75.52 | 68.26 |
BERT-Soft-Lexicon(LSTM) | 67.01 | 76.66 | 71.51 |
Ours | 68.34 | 81.56 | 74.37 |