Skip to main content

Table 5 Performance of different kernel methods in five types of corpuses

From: Multiple kernels learning-based biological entity relationship extraction method

Corpus set

Evaluation parameters

Characteristics- based kernels

Extension path dependency kernel

Tag graph kernel

Kernels from

     

three-kernel fusion

Aimed

P

45.34

42.31

50.82

57.45

 

R

61.25

68.54

69.76

72.31

 

F

55.36

52.63

58.61

60.98

IEPA

P

56.84

52.48

61.58

73.82

 

R

72.92

69.35

76.91

81.06

 

F

87.15

63.79

74.23

79.57

BioInfer

P

45.61

41.32

51.64

91.69

 

R

63.57

69.76

68.92

71.62

 

F

56.24

52.58

59.73

62.35

HPRD

P

50.26

49.96

55.64

61.87

 

R

67.59

66.31

67.81

72.35

 

F

75.38

69.78

70.01

85.48

LLL

P

53.59

83.34

71.92

75.69

 

R

70.12

69.78

70.84

78.37

 

F

68.43

88.03

77.43

90.12