RTF: Region-based Table Filling Method for Relational Triple Extraction

TL;DR

This paper introduces RTF, a novel region-based table filling method for relational triple extraction that captures local spatial dependencies and improves boundary detection, leading to state-of-the-art results on benchmark datasets.

Contribution

The paper proposes a new region-based tagging scheme and bi-directional decoding strategy, enhancing entity boundary detection and relation classification in knowledge graph construction.

Findings

Achieves state-of-the-art performance on benchmark datasets.

Improves entity boundary detection by modeling local spatial dependencies.

Enhances relation classification efficiency through shared tagging scores.

Abstract

Relational triple extraction is crucial work for the automatic construction of knowledge graphs. Existing methods only construct shallow representations from a token or token pair-level. However, previous works ignore local spatial dependencies of relational triples, resulting in a weakness of entity pair boundary detection. To tackle this problem, we propose a novel Region-based Table Filling method (RTF). We devise a novel region-based tagging scheme and bi-directional decoding strategy, which regard each relational triple as a region on the relation-specific table, and identifies triples by determining two endpoints of each region. We also introduce convolution to construct region-level table representations from a spatial perspective which makes triples easier to be captured. In addition, we share partial tagging scores among different relations to improve learning efficiency of relation classifier. Experimental results show that our method achieves state-of-the-art with better generalization capability on three variants of two widely used benchmark datasets.