Assistant diagnosis with Chinese electronic medical records based on CNN and BiLSTM with phrase-level and word-level attentions

Table 3 Prediction result for each disease of FCNBLA and its baselines on the test set

F-score (%)	FCNBLA	SVM_TFIDF	CNN	RCNN	BiLSTM
Diseases	FCNBLA	SVM_TFIDF	CNN	RCNN	BiLSTM
diabetes	96.29	95.27	96.08	96.22	95.93
hypertension	89.24	87.46	88.72	89.37	89.30
COPD	96.72	96.57	96.64	96.66	96.49
arrhythmia	87.68	84.88	87.35	87.23	86.35
asthma	93.59	92.06	92.81	94.56	93.43
gastritis	83.25	74.82	81.36	79.90	79.12
gastric polyps	89.90	84.77	88.74	88.37	88.82
gout	91.89	90.35	90.24	91.76	90.15
gastric ulcers	90.48	84.75	89.39	89.94	89.66
UTI	93.91	89.29	91.08	91.38	91.74

ISSN: 1471-2105