Table 5 The influence of \(\alpha\) and \(\beta\) on VADGAT’s performance.

From: Triple dimensional psychology knowledge encouraging graph attention networks to exploit aspect-based sentiment analysis

\(\alpha\) &\(\beta\)

Lap14

Twitter

Rest14

Rest15

Rest16

Acc(%)

F1(%)

Acc(%)

F1(%)

Acc(%)

F1(%)

Acc(%)

F1(%)

Acc(%)

F1(%)

0.1 &0.9

83.54

81.08

76.88

75.82

87.14

81.48

85.79

72.26

93.18

77.85

0.2 &0.8

82.45

79.51

77.17

76.14

86.61

80.42

86.16

74.30

93.18

77.65

0.3 &0.7

82.76

79.81

77.17

76.30

87.32

81.09

85.61

72.45

91.72

77.44

0.4 &0.6

81.66

78.38

77.31

76.21

86.16

78.93

86.53

75.48

91.72

78.25

0.5 &0.5

82.29

78.93

76.45

74.88

87.59

81.33

86.16

72.04

91.88

77.70

0.6 &0.4

81.66

78.81

77.17

75.98

87.05

80.44

87.45

74.09

91.56

79.09

0.7 &0.3

83.07

80.10

78.18

76.92

87.50

81.11

88.93

73.42

93.18

81.51

0.8 &0.2

81.82

78.79

77.60

76.70

87.14

80.79

87.08

72.38

91.40

79.08

0.9 &0.1

82.92

80.32

77.60

76.35

86.88

80.44

88.19

75.31

92.37

79.49

  1. The best result is in bold, and the suboptimal result is marked by underline.