Skip to main content

Table 3 Performance of different models

From: B-LBConA: a medical entity disambiguation model based on Bio-LinkBERT and context-aware mechanism

Model

NCBI

ADR

ShARe/CLEF

test

refined test

test

refined test

test

refined test

BERT-based Ranking [11]

88.96

67.44

93.17

79.83

91.09

80.47

Edge-weight-updating NN [34]

91.72

71.15

92.21

80.05

91.56

81.45

SciFive [35]

90.47

69.53

92.17

75.18

91.01

79.83

ED-GNN(GraphSAGE) [1]

92.44

72.36

92.03

78.25

89.46

76.39

D-C + OD-T [36]

92.25

-

-

-

90.41

-

ResCNN [37]

92.40

73.02

\(\underline{93.83}\)

78.96

92.79

79.13

Lightweight-NN [19]

92.56

69.65

93.07

80.34

92.73

80.78

KRISSBERT [38]

89.93

70.88

91.65

75.42

90.41

78.92

Inter- and Intra-Attention [13]

91.28

-

93.13

-

-

-

G-MAP [39]

\(\underline{92.61}\)

\(\underline{73.75}\)

93.26

79.23

\(\underline{92.98}\)

\(\underline{81.29}\)

B-LBConA (our model)

93.57

74.38

94.72

\(\underline{79.89}\)

94.23

80.68

  1. The best performance on each dataset is marked in bold, and the second-best performance is marked in underline; “−” means the result is not provided