ER: Equivariance Regularizer for Knowledge Graph Completion

TL;DR

This paper introduces the Equivariance Regularizer (ER), a novel method that leverages semantic equivariance to improve generalization and reduce overfitting in knowledge graph completion models.

Contribution

The paper proposes a new regularizer, ER, that exploits semantic equivariance, enhancing model generalization across various KGC models and outperforming state-of-the-art methods.

Findings

ER significantly improves relation prediction accuracy.

It effectively suppresses overfitting in KGC models.

The approach is applicable to multiple model types.

Abstract

Tensor factorization and distanced based models play important roles in knowledge graph completion (KGC). However, the relational matrices in KGC methods often induce a high model complexity, bearing a high risk of overfitting. As a remedy, researchers propose a variety of different regularizers such as the tensor nuclear norm regularizer. Our motivation is based on the observation that the previous work only focuses on the "size" of the parametric space, while leaving the implicit semantic information widely untouched. To address this issue, we propose a new regularizer, namely, Equivariance Regularizer (ER), which can suppress overfitting by leveraging the implicit semantic information. Specifically, ER can enhance the generalization ability of the model by employing the semantic equivariance between the head and tail entities. Moreover, it is a generic solution for both distance based models and tensor factorization based models. The experimental results indicate a clear and substantial improvement over the state-of-the-art relation prediction methods.