TransA: An Adaptive Approach for Knowledge Graph Embedding

TL;DR

TransA introduces an adaptive metric approach for knowledge graph embedding, enhancing the flexibility and accuracy of representing complex entities and relations, and outperforms existing methods on benchmark datasets.

Contribution

It proposes TransA, a novel adaptive metric learning method that improves knowledge graph embedding by addressing limitations of translation-based models.

Findings

Significant improvements over state-of-the-art baselines

Consistent performance gains on benchmark datasets

Enhanced modeling of complex entities and relations

Abstract

Knowledge representation is a major topic in AI, and many studies attempt to represent entities and relations of knowledge base in a continuous vector space. Among these attempts, translation-based methods build entity and relation vectors by minimizing the translation loss from a head entity to a tail one. In spite of the success of these methods, translation-based methods also suffer from the oversimplified loss metric, and are not competitive enough to model various and complex entities/relations in knowledge bases. To address this issue, we propose \textbf{TransA}, an adaptive metric approach for embedding, utilizing the metric learning ideas to provide a more flexible embedding method. Experiments are conducted on the benchmark datasets and our proposed method makes significant and consistent improvements over the state-of-the-art baselines.