Research Article
Ptr4BERT: Automatic Semisupervised Chinese Government Message Text Classification Method Based on Transformer-Based Pointer Generator Network
Table 5
Performance comparison of different methods in English dataset.
| Method | govMes1 | govMes2 | P | R | F1 | P | R | F1 |
| SVM | 81.26 | 80.21 | 80.73 | 79.25 | 78.32 | 78.78 | TextRNN | 86.11 | 87.23 | 86.66 | 86.12 | 85.36 | 85.74 | TextCNN | 85.66 | 85.27 | 85.46 | 84.07 | 85.72 | 84.89 | BERT | 89.26 | 88.59 | 88.92 | 87.21 | 88.04 | 87.62 | Ptr4BERT (Ours) | 90.83 | 89.57 | 90.20 | 90.12 | 89.86 | 89.99 |
|
|