Skip to main content

Table 2 Comparison results of SEoG and other different methods on the CDR and CHR test sets (%)

From: Document-level medical relation extraction via edge-oriented graph neural network based on document structure and external knowledge

Dataset

Method

Overall

Intra

Inter

P

R

F1

F1

F1

CDR

Gu et al. [3]

55.7

68.1

61.3

57.2

11.7

Verga et al. [20]

55.6

70.8

62.1

–

–

Nguyen and Verspoor [5]

57.0

68.6

62.3

–

–

Sahu et al. [9]

52.8

66.0

58.6

–

–

Christopoulou et al. [10]

62.1

65.2

63.6

68.2

50.9

KEoG (node) [11]

65.4

71.2

68.2

71.8

58.3

SEoG

64.5

75.5

69.6

73.4

59.9

CHR

CNN-RE [9]

81.2

87.3

84.1

–

–

RNN-RE [9]

83.0

90.1

86.4

–

–

Sahu et al. [9]

84.7

90.5

87.5

–

–

KEoG (node) [11]

89.9

92.6

91.2

93.4

86.3

SEoG

88.3

92.6

90.4

93.1

84.4

  1. Bold highlight the highest result on a given dataset in our experiments