Table 6 Results of ablation experiments comparing different feature extraction models.
From: Transformer-based prototype network for Chinese nested named entity recognition
F1 | ACE2005 | ChiNesE | People’s Daily |
|---|---|---|---|
BERT | 86.15 | 84.82 | 96.32 |
roberta | 86.51 | 84.20 | 96.03 |
longformer | 85.71 | 84.07 | 96.07 |
roformer | 81.45 | 76.68 | 92.71 |
-LSTM | 84.48 | 84.20 | 95.41 |
