Prior Bilinear Based Models for Knowledge Graph Completion

TL;DR

This paper introduces the UniBi model, a bilinear approach for knowledge graph completion that captures prior properties like the law of identity, improving interpretability and performance over existing models.

Contribution

The paper proposes the UniBi model, addressing the lack of prior property modeling in bilinear models, and demonstrates its theoretical and practical advantages.

Findings

UniBi models the prior property effectively.

Enhanced interpretability of the model.

Improved performance in knowledge graph completion tasks.

Abstract

Bilinear based models are powerful and widely used approaches for Knowledge Graphs Completion (KGC). Although bilinear based models have achieved significant advances, these studies mainly concentrate on posterior properties (based on evidence, e.g. symmetry pattern) while neglecting the prior properties. In this paper, we find a prior property named "the law of identity" that cannot be captured by bilinear based models, which hinders them from comprehensively modeling the characteristics of KGs. To address this issue, we introduce a solution called Unit Ball Bilinear Model (UniBi). This model not only achieves theoretical superiority but also offers enhanced interpretability and performance by minimizing ineffective learning through minimal constraints. Experiments demonstrate that UniBi models the prior property and verify its interpretability and performance.