Skip to main content

Table 4 Results of the document subgraph with different sizes of the sliding window for training and testing

From: Exploiting document graphs for inter sentence relation extraction

w for training

w for testing

Precision

Recall

F1

1

1

65.34

55.50

60.02

 

2

62.20

57.22

59.61

 

3

61.47

58.27

59.83

 

4

61.92

54.86

58.18

 

5

57.13

59.76

58.42

2

1

61.95

60.19

61.06

 

2

61.25

61.26

61.25

 

3

61.97

60.30

61.12

 

4

61.30

58.52

59.88

 

5

60.99

59.36

60.16

3

1

61.05

61.74

61.39

 

2

60.65

61.74

61.19

 

3

60.70

61.88

61.28

 

4

62.30

59.47

60.85

 

5

61.10

59.81

60.45

4

1

60.30

64.01

62.10

 

2

57.88

65.98

61.67

 

3

58.31

65.27

61.59

 

4

58.40

63.86

61.01

 

5

59.97

61.71

60.83

5

1

61.15

63.76

62.43

 

2

60.13

65.89

62.88

 

3

58.56

65.79

61.96

 

4

58.64

62.42

60.47

 

5

57.92

62.36

60.06

  1. Results are reported in %
  2. The highest result in each column is highlighted in bold