Skip to main content

Table 4 Evaluation results of the four systems according to the two annotators

From: Wide-coverage relation extraction from MEDLINE using deep syntax

System

Annotator A

Annotator B

Mean

 

Pre.

Re.

F.

Pre.

Re.

F.

Pre.

Re.

F.

ReVerb

44.15

6.75

11.72

61.04

9.34

16.20

52.59

8.05

13.96

OLLIE

40.85

13.32

20.10

53.65

17.49

26.38

47.25

15.41

23.24

SemRep

59.37

40.95

48.47

65.13

38.83

48.65

62.25

39.89

48.56

PASMED

43.27

67.19

52.65

51.50

69.24

59.13

47.39

68.22

55.89

  1. SemRep achieves the highest precision, PASMED achieves the highest relative recall.