| 模型 | Accuracy | Precision | 加权F1值 |
| 朴素贝叶斯 | 62.09% | 64.47% | 0.6295 |
| Logistic回归 | 64.23% | 65.79% | 0.6485 |
| BERT(base, cased) | 68.11% | 69.28% | 0.6943 |
| BERT(base, large) | 69.13% | 69.87% | 0.6958 |
| DistilBERT(base, cased) | 67.77% | 69.43% | 0.6901 |
| RoBERTa(base) | 69.75% | 70.94% | 0.6905 |