Table 3 Comparative experiment results table (FewFC).
Model | Event Type Classification | Trigger Extraction | Argument Extraction | Mean F1 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
P | R | F1 | P | R | F1 | P | R | F1 | ||
BERT | 89.92 | 79.00 | 84.00 | 80.17 | 61.82 | 69.82 | 74.63 | 62.73 | 68.16 | 73.99 |
BERT-CRF-Joint | 90.83 | 79.78 | 84.43 | 80.72 | 63.00 | 70.81 | 74.23 | 61.18 | 67.14 | 74.13 |
CasEE | 87.50 | 87.73 | 88.64 | 77.90 | 78.53 | 82.88 | 76.83 | 77.12 | 76.91 | 82.81 |
GPT2 | 82.08 | 83.30 | 82.69 | 82.35 | 83.41 | 82.88 | 76.38 | 77.62 | 76.99 | 80.85 |
PosEKE-GPT2(ours) | 91.88 | 93.10 | 92.48 | 91.18 | 92.24 | 91.70 | 81.74 | 82.99 | 82.36 | 88.85 |
