CORE: Cyclic Orthotope Relation Embedding for Knowledge Graph Completion

TL;DR

CORE introduces a boundary-less torus manifold embedding for knowledge graphs, enabling seamless relation region wrapping and adaptive regularization to improve link prediction accuracy.

Contribution

It proposes a novel cyclic orthotope relation embedding model on a torus manifold, addressing boundary constraints and capturing complex relation patterns.

Findings

CORE achieves state-of-the-art performance on benchmark datasets.

The model effectively captures complex logical relation patterns.

Experimental results show significant improvement in link prediction accuracy.

Abstract

Knowledge graph completion (KGC) aims to automatically infer missing facts in multi-relational data by mapping entities and relations into continuous representation spaces. Recent region-based embedding models have shown great promise in capturing complex logical patterns by representing relations as geometric regions. However, these models inevitably suffer from absolute boundary constraints during optimization. Conversely, without such constraints, relation regions expand indefinitely. To address the limitation, we propose \textbf{CORE} (Cyclic Orthotope Relation Embedding), a novel KGC model that embeds entities and relations onto a boundary-less torus manifold.CORE represents relations as cyclic orthotopes on the torus manifold, allowing regions to seamlessly wrap around spatial boundaries to ensure smooth gradient conduction. Furthermore, an adaptive width regularization is introduced to prevent unconditional region expansion. Theoretical analysis proves that CORE can capture various complex relation patterns such as subsumption and intersection. Extensive experiments on four benchmark datasets demonstrate that CORE achieves highly competitive performance, significantly improving link prediction accuracy in dense semantic environments.