Table 4 Comparison of CapsTM and other-state-of-the-art methods
From: CapsTM: capsule network for Chinese medical text matching
Model | F-score | Precision | Recall |
|---|---|---|---|
ESIM | 0.8283 | 0.8156 | 0.8413 |
ABCNN | 0.8144 | 0.8165 | 0.8123 |
BIMPM | 0.8367 | 0.8190 | 0.8555 |
DISAN | 0.8326 | 0.7800 | 0.8930 |
DRCN | 0.8047 | 0.7877 | 0.8224 |
DECOMP | 0.7951 | 0.7410 | 0.8577 |
CapsTM (word2vec) | 0.8432 | 0.8364 | 0.8501 |
BERT | 0.8646 | 0.8603 | 0.8689 |
CapsTM (BERT) | 0.8666 | 0.8655 | 0.8677 |