Table 7 Accuracy evaluation metrics results on Reddit dataset.

From: A dynamic graph structural framework for implicit sentiment identification based on complementary semantic and structural information

Methods

Body (%)

Punchline (%)

Full (%)

DPCNN

64.63

66.77

67.10

HAN

63.81

66.44

69.40

TEXTCNN

62.17

67.26

66.28

RCNN

64.47

69.07

68.75

IDGL

61.18

63.15

65.95

KIG*

64.37

64.25

67.84

SIF

65.36

70.28

71.55

  1. The star symbol indicates that we cite results from the original paper.
  2. Best values are in bold.
Back to article page