Rethinking Graph Convolutional Networks in Knowledge Graph Completion

TL;DR

This paper critically examines the role of GCNs in knowledge graph completion, revealing that graph structure modeling is less impactful than transformations of entity representations, leading to a simpler, more efficient framework LTE-KGE.

Contribution

The paper demonstrates that GCNs' graph modeling does not significantly improve KGC and introduces LTE-KGE, a simple framework using linear transformations of entity embeddings that matches GCN performance with less complexity.

Findings

Graph structure modeling in GCNs has minimal impact on KGC performance.

Transformations of entity representations are key to performance improvements.

LTE-KGE achieves similar results to GCN-based models with higher efficiency.

Abstract

Graph convolutional networks (GCNs) -- which are effective in modeling graph structures -- have been increasingly popular in knowledge graph completion (KGC). GCN-based KGC models first use GCNs to generate expressive entity representations and then use knowledge graph embedding (KGE) models to capture the interactions among entities and relations. However, many GCN-based KGC models fail to outperform state-of-the-art KGE models though introducing additional computational complexity. This phenomenon motivates us to explore the real effect of GCNs in KGC. Therefore, in this paper, we build upon representative GCN-based KGC models and introduce variants to find which factor of GCNs is critical in KGC. Surprisingly, we observe from experiments that the graph structure modeling in GCNs does not have a significant impact on the performance of KGC models, which is in contrast to the common belief. Instead, the transformations for entity representations are responsible for the performance improvements. Based on the observation, we propose a simple yet effective framework named LTE-KGE, which equips existing KGE models with linearly transformed entity embeddings. Experiments demonstrate that LTE-KGE models lead to similar performance improvements with GCN-based KGC methods, while being more computationally efficient. These results suggest that existing GCNs are unnecessary for KGC, and novel GCN-based KGC models should count on more ablation studies to validate their effectiveness. The code of all the experiments is available on GitHub at https://github.com/MIRALab-USTC/GCN4KGC.