Table 4 Performance of CSGAT and Two Variant Models: One variant removes attribute context, while the other removes all context information.

From: Contextual semantics graph attention network model for entity resolution

Models

Amazon-Google

BeerAdvo-RateBeer

F1 (%)

ΔF1 (%)

F1 (%)

ΔF1 (%)

CSGAT

65.88

_

82.73

_

CSGATnon-Attribute

62.45

-5.21

75.33

-9.82

CSGATnon-Context

58.68

-10.92

68.41

-17.31

  1. Through experimental comparison, the semantic gains of these two types of context information are obtained.
  2. Significant values are in [bold].
Back to article page