| 模型 | NLPCC 2016CKBQA | CCKS 2019EL | ||||
| Top 1 | Top 2 | Top 3 | Top 1 | Top 2 | Top 3 | |
| State-of-the-Art | ||||||
| BB-KBQA [7] | 86.27 | 90.18 | 91.97 | - | - | - |
| 方法 [8] | - | - | - | 80.16 | 84.56 | 85.2 |
| 知识增强的预训练模型 | ||||||
| K-BERT [19] | 86.52 | 89.83 | 91.72 | 78.73 | 82.76 | 83.48 |
| EKBERT | ||||||
| EKBERT | 92.92 | 96.31 | 97.56 | 80.32 | 85.03 | 85.78 |