Skip to main content

Table 6 Effects of different parameter sharing strategies on performance on the CDR dataset

From: Knowledge-guided convolutional networks for chemical-disease relation extraction

Method

Intra-sentence level

Inter-sentence level

Document level

P (%)

R (%)

F (%)

P (%)

R (%)

F (%)

P (%)

R (%)

F (%)

KCN

70.61

60.41

65.12

65.37

12.57

21.09

69.65

72.98

71.28

SGate-SAtt

70.09

60.23

64.78

62.96

9.57

16.61††

69.02

69.79

69.40

DGate-DAtt

70.39

60.23

64.91

68.13

10.23

17.78††

70.06

70.45

70.25††

SGate-DAtt

69.15

60.98

64.81

63.19

10.79

18.43††

68.18

71.76

69.93††

  1. The descriptions and analysis for Table 6 could be found in subsection “Effects of sharing parameters”. The marker and †† represent P-value < 0.05 and P-value < 0.01, respectively, using pairwise t-test against KCN. The highest scores are highlighted in bold