Contextualised Graph Attention for Improved Relation Extraction

TL;DR

This paper introduces a novel contextualized graph attention network that leverages multiple sub-graphs and edge features to enhance relation extraction, achieving state-of-the-art results on a benchmark dataset.

Contribution

It proposes a new method combining multiple sub-graphs and edge features within graph attention networks for improved relation extraction.

Findings

Achieves an F1-score of 86.3 on Semeval 2010 Task 8.

Effectively combines edge features with GAT and GCN models.

Demonstrates state-of-the-art performance in relation extraction.

Abstract

This paper presents a contextualized graph attention network that combines edge features and multiple sub-graphs for improving relation extraction. A novel method is proposed to use multiple sub-graphs to learn rich node representations in graph-based networks. To this end multiple sub-graphs are obtained from a single dependency tree. Two types of edge features are proposed, which are effectively combined with GAT and GCN models to apply for relation extraction. The proposed model achieves state-of-the-art performance on Semeval 2010 Task 8 dataset, achieving an F1-score of 86.3.