Skip to main content

Table 7 Evaluation metrics obtained with spaCy model for each named entity

From: De-identifying Spanish medical texts - named entity recognition applied to radiology reports

 

Training set

Validation set

Test set

MEDDOCAN

 

Precision

Recall

F1

Precision

Recall

F1

Precision

Recall

F1

Precision

Recall

F1

CAB

99.43

96.54

97.96

98.28

93.98

96.08

92.54

74.49

82.52

4.76

33.33

8.33

DIR

100

100

100

94.28

63.96

76.01

87.79

74.77

61.46

43.15

4.47

8.01

FECHA

100

100

100

98.54

99.04

98.78

98.20

97.53

97.86

51.39

89.41

65.13

INST

99.97

99.96

99.98

98.19

97.24

97.71

93.50

98.00

95.69

45.72

12.28

19.27

LOC

100

100

100

76.64

54.66

63.80

61.04

26.85

36.79

7.19

0.32

0.59

NAME

100

99.99

99.99

98.34

98.28

98.31

88.78

94.29

93.19

75.62

83.91

79.23

NUM

100

100

100

97.81

95.65

96.72

95.11

87.56

91.18

68.50

60.32

63.99

 

99.87

99.28

99.58

98.06

96.10

97.08

93.23

89.39

91.31

65.63

55.37

59.98