Skip to main content

Table 6 Comparison results of NER on discharge summaries

From: A multitask bi-directional RNN model for named entity recognition on Chinese electronic medical records

 

Bi-RNN model

Our proposed model

Entity type

Precision

Recall

F-measure

Precision

Recall

F-measure

Disease

82.82

78.02

80.34

84.11

84.70

84.40

Symptom

80.26

80.11

80.19

88.08

84.01

86.00

Disease group

37.50

100

54.54

43.75

82.35

57.14

Treatment

68.89

78.58

73.41

73.91

82.06

77.77

Test

82.99

86.43

84.68

89.23

87.99

88.61

Macro average

70.91

84.67

74.63

75.82

84.22

78.79