Chinese Word Segmentation On Msr
평가 지표
F1
Precision
Recall
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | F1 | Precision | Recall | Paper Title | Repository |
|---|---|---|---|---|---|
| Glyce + BERT | 98.3 | 98.2 | 98.3 | Glyce: Glyph-vectors for Chinese Character Representations | |
| BABERT-LE | 98.63 | - | - | Unsupervised Boundary-Aware Language Model Pretraining for Chinese Sequence Labeling | |
| ZEN (Random Init) | 97.89 | - | - | ZEN: Pre-training Chinese Text Encoder Enhanced by N-gram Representations | |
| BABERT | 98.44 | - | - | Unsupervised Boundary-Aware Language Model Pretraining for Chinese Sequence Labeling | |
| WMSeg + ZEN | 98.40 | - | - | Improving Chinese Word Segmentation with Wordhood Memory Networks | |
| ZEN (Init with Chinese BERT) | 98.35 | - | - | ZEN: Pre-training Chinese Text Encoder Enhanced by N-gram Representations |
0 of 6 row(s) selected.