Research Article
Recognition of Chinese Legal Elements Based on Transfer Learning and Semantic Relevance
Table 7
Comparative results of different methods.
| | Model | Divorce | Labor | Loan | | MiF1 | MaF1 | MiF1 | MaF1 | MiF1 | MaF1 |
| | TextCNN | 78.7% | 65.9% | 76.4% | 54.4% | 80.3% | 60.6% | | DPCNN | 81.3% | 64.0% | 79.8% | 47.4% | 81.4% | 42.5% | | LSTM | 80.6% | 67.3% | 81.0% | 52.9% | 80.4% | 53.1% | | BERT-SS | 83.3% | 69.1% | 76.8% | 43.7% | 78.6% | 39.5% | | BERT-SP | 82.8% | 67.6% | 79.2% | 47.1% | 80.0% | 43.7% | | BERT-LER | 83.1% | 69.6% | 80.9% | 54.6% | 81.0% | 56.5% |
|
|
