Similarity matching of medical question based on Siamese network

Table 6 Effectiveness of different models on the ethnic medicine question datasets (unit %)

Model	Precision	Recall	F1-score
ABCNN	86.55	83.68	85.09
ESIM	87.72	90.19	88.94
BIMPM	90.70	90.26	90.48
Siamese-BiLSTM	89.35	92.17	90.74
Siamese-BiGRU	91.49	93.38	92.43
Siamese-BiLSTM-attention	92.97	93.87	93.42
Siamese-BiGRU-attention	93.05	96.78	94.88
Siamese-BiLSTM-Attention-CNN	95.22	94.86	95.04
Siamese-BiGRU-Attention-CNN	97.15	98.82	97.98

ISSN: 1472-6947