Table 6 Open in new tab Comparison of...

Model^a	Performance
Bi-Bi-S	91.15 ± 0.21	79.04 ± 0.43	85.92 ± 0.18
Bi-I-S	91.29 ± 0.29	78.36 ± 0.32	86.19 ± 0.34
Bi-BE-S	91.69 ± 0.24	79.75 ± 0.29	86.18 ± 0.42
Bi-Bi-T	91.81 ± 0.16	80.26 ± 0.51	85.65 ± 0.31
Bi-I-T	92.91 ± 0.33	83.07 ± 0.38	87.71 ± 0.28
Bi-BE-T	92.31 ± 0.22	81.68 ± 0.30	86.67 ± 0.32
I-Bi-S	91.23 ± 0.32	78.86 ± 0.33	85.11 ± 0.29
I-I-S	91.40 ± 0.19	84.70 ± 0.42	89.52 ± 0.51
I-BE-S	89.64 ± 0.28	75.82 ± 0.22	82.16 ± 0.27
I-Bi-T	92.28 ± 0.30	81.49 ± 0.21	87.86 ± 0.25
I-I-T	91.87 ± 0.26	83.09 ± 0.37	87.26 ± 0.17
I-BE-T	92.28 ± 0.41	86.91 ± 0.33	90.26 ± 0.26
BE-Bi-S	91.19 ± 0.22	81.22 ± 0.11	86.83 ± 0.15
BE-I-S	92.28 ± 0.18	81.54 ± 0.51	86.58 ± 0.26
BE-BE-S	92.02 ± 0.36	85.84 ± 0.62	89.75 ± 0.41
BE-Bi-T	92.41 ± 0.31	79.16 ± 0.37	85.27 ± 0.33
BE-I-T	92.13 ± 0.28	83.21 ± 0.28	87.87 ± 0.19
BE-BE-T	93.76 ± 0.27	85.50 ± 0.18	89.44 ± 0.22
V-V-V^b	93.14 ± 0.18	89.20 ± 0.40	91.13 ± 0.21

Model^a

Performance

Precision (%)

Recall (%)

F1 score (%)

Bi-Bi-S

91.15 ± 0.21

79.04 ± 0.43

85.92 ± 0.18

Bi-I-S

91.29 ± 0.29

78.36 ± 0.32

86.19 ± 0.34

Bi-BE-S

91.69 ± 0.24

79.75 ± 0.29

86.18 ± 0.42

Bi-Bi-T

91.81 ± 0.16

80.26 ± 0.51

85.65 ± 0.31

Bi-I-T

92.91 ± 0.33

83.07 ± 0.38

87.71 ± 0.28

Bi-BE-T

92.31 ± 0.22

81.68 ± 0.30

86.67 ± 0.32

I-Bi-S

91.23 ± 0.32

78.86 ± 0.33

85.11 ± 0.29

I-I-S

91.40 ± 0.19

84.70 ± 0.42

89.52 ± 0.51

I-BE-S

89.64 ± 0.28

75.82 ± 0.22

82.16 ± 0.27

I-Bi-T

92.28 ± 0.30

81.49 ± 0.21

87.86 ± 0.25

I-I-T

91.87 ± 0.26

83.09 ± 0.37

87.26 ± 0.17

I-BE-T

92.28 ± 0.41

86.91 ± 0.33

90.26 ± 0.26

BE-Bi-S

91.19 ± 0.22

81.22 ± 0.11

86.83 ± 0.15

BE-I-S

92.28 ± 0.18

81.54 ± 0.51

86.58 ± 0.26

BE-BE-S

92.02 ± 0.36

85.84 ± 0.62

89.75 ± 0.41

BE-Bi-T

92.41 ± 0.31

79.16 ± 0.37

85.27 ± 0.33

BE-I-T

92.13 ± 0.28

83.21 ± 0.28

87.87 ± 0.19

BE-BE-T

93.76 ± 0.27

85.50 ± 0.18

89.44 ± 0.22

V-V-V^b

93.14 ± 0.18

89.20 ± 0.40

91.13 ± 0.21

Model^a	Performance
Bi-Bi-S	91.15 ± 0.21	79.04 ± 0.43	85.92 ± 0.18
Bi-I-S	91.29 ± 0.29	78.36 ± 0.32	86.19 ± 0.34
Bi-BE-S	91.69 ± 0.24	79.75 ± 0.29	86.18 ± 0.42
Bi-Bi-T	91.81 ± 0.16	80.26 ± 0.51	85.65 ± 0.31
Bi-I-T	92.91 ± 0.33	83.07 ± 0.38	87.71 ± 0.28
Bi-BE-T	92.31 ± 0.22	81.68 ± 0.30	86.67 ± 0.32
I-Bi-S	91.23 ± 0.32	78.86 ± 0.33	85.11 ± 0.29
I-I-S	91.40 ± 0.19	84.70 ± 0.42	89.52 ± 0.51
I-BE-S	89.64 ± 0.28	75.82 ± 0.22	82.16 ± 0.27
I-Bi-T	92.28 ± 0.30	81.49 ± 0.21	87.86 ± 0.25
I-I-T	91.87 ± 0.26	83.09 ± 0.37	87.26 ± 0.17
I-BE-T	92.28 ± 0.41	86.91 ± 0.33	90.26 ± 0.26
BE-Bi-S	91.19 ± 0.22	81.22 ± 0.11	86.83 ± 0.15
BE-I-S	92.28 ± 0.18	81.54 ± 0.51	86.58 ± 0.26
BE-BE-S	92.02 ± 0.36	85.84 ± 0.62	89.75 ± 0.41
BE-Bi-T	92.41 ± 0.31	79.16 ± 0.37	85.27 ± 0.33
BE-I-T	92.13 ± 0.28	83.21 ± 0.28	87.87 ± 0.19
BE-BE-T	93.76 ± 0.27	85.50 ± 0.18	89.44 ± 0.22
V-V-V^b	93.14 ± 0.18	89.20 ± 0.40	91.13 ± 0.21

Table 6

Open in new tab

Comparison of end-to-end identification effect

Model^a	Performance
Model^a	Precision (%)	Recall (%)	F1 score (%)
Bi-Bi-S	91.15 ± 0.21	79.04 ± 0.43	85.92 ± 0.18
Bi-I-S	91.29 ± 0.29	78.36 ± 0.32	86.19 ± 0.34
Bi-BE-S	91.69 ± 0.24	79.75 ± 0.29	86.18 ± 0.42
Bi-Bi-T	91.81 ± 0.16	80.26 ± 0.51	85.65 ± 0.31
Bi-I-T	92.91 ± 0.33	83.07 ± 0.38	87.71 ± 0.28
Bi-BE-T	92.31 ± 0.22	81.68 ± 0.30	86.67 ± 0.32
I-Bi-S	91.23 ± 0.32	78.86 ± 0.33	85.11 ± 0.29
I-I-S	91.40 ± 0.19	84.70 ± 0.42	89.52 ± 0.51
I-BE-S	89.64 ± 0.28	75.82 ± 0.22	82.16 ± 0.27
I-Bi-T	92.28 ± 0.30	81.49 ± 0.21	87.86 ± 0.25
I-I-T	91.87 ± 0.26	83.09 ± 0.37	87.26 ± 0.17
I-BE-T	92.28 ± 0.41	86.91 ± 0.33	90.26 ± 0.26
BE-Bi-S	91.19 ± 0.22	81.22 ± 0.11	86.83 ± 0.15
BE-I-S	92.28 ± 0.18	81.54 ± 0.51	86.58 ± 0.26
BE-BE-S	92.02 ± 0.36	85.84 ± 0.62	89.75 ± 0.41
BE-Bi-T	92.41 ± 0.31	79.16 ± 0.37	85.27 ± 0.33
BE-I-T	92.13 ± 0.28	83.21 ± 0.28	87.87 ± 0.19
BE-BE-T	93.76 ± 0.27	85.50 ± 0.18	89.44 ± 0.22
V-V-V^b	93.14 ± 0.18	89.20 ± 0.40	91.13 ± 0.21

Model^a	Performance
Model^a	Precision (%)	Recall (%)	F1 score (%)
Bi-Bi-S	91.15 ± 0.21	79.04 ± 0.43	85.92 ± 0.18
Bi-I-S	91.29 ± 0.29	78.36 ± 0.32	86.19 ± 0.34
Bi-BE-S	91.69 ± 0.24	79.75 ± 0.29	86.18 ± 0.42
Bi-Bi-T	91.81 ± 0.16	80.26 ± 0.51	85.65 ± 0.31
Bi-I-T	92.91 ± 0.33	83.07 ± 0.38	87.71 ± 0.28
Bi-BE-T	92.31 ± 0.22	81.68 ± 0.30	86.67 ± 0.32
I-Bi-S	91.23 ± 0.32	78.86 ± 0.33	85.11 ± 0.29
I-I-S	91.40 ± 0.19	84.70 ± 0.42	89.52 ± 0.51
I-BE-S	89.64 ± 0.28	75.82 ± 0.22	82.16 ± 0.27
I-Bi-T	92.28 ± 0.30	81.49 ± 0.21	87.86 ± 0.25
I-I-T	91.87 ± 0.26	83.09 ± 0.37	87.26 ± 0.17
I-BE-T	92.28 ± 0.41	86.91 ± 0.33	90.26 ± 0.26
BE-Bi-S	91.19 ± 0.22	81.22 ± 0.11	86.83 ± 0.15
BE-I-S	92.28 ± 0.18	81.54 ± 0.51	86.58 ± 0.26
BE-BE-S	92.02 ± 0.36	85.84 ± 0.62	89.75 ± 0.41
BE-Bi-T	92.41 ± 0.31	79.16 ± 0.37	85.27 ± 0.33
BE-I-T	92.13 ± 0.28	83.21 ± 0.28	87.87 ± 0.19
BE-BE-T	93.76 ± 0.27	85.50 ± 0.18	89.44 ± 0.22
V-V-V^b	93.14 ± 0.18	89.20 ± 0.40	91.13 ± 0.21

^aBi, BE, I, T and S represented BiLSTM, BERT, the IDCNN, the Transformer and SEQ2SEQ, respectively.

^bThe first V indicates voting for the binary classification model (that is, the recognition results that are the same between two of the three models shall prevail); the second V indicates voting for the entity recognition model (that is, the intersection of the entity recognition results of the three models is taken and deduplicated); the third V indicates voting for the entity alignment model (that is, the intersection of the alignment results of the two models is taken and deduplicated).

This Feature Is Available To Subscribers Only