Contextualised Graph Attention for Improved Relation Extraction

2020/04/22 by Mandya, Angrosh, Bollegala, Danushka, Coenen, Frans
#Computation and Language (cs.CL) #FOS: Computer and information sciences #Information Retrieval (cs.IR)

paper · doi:10.48550/arxiv.2004.10624

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.

Related

• Attention Guided Graph Convolutional Networks for Relation Extraction
• r-GAT: Relational Graph Attention Network for Multi-Relational Graphs
• Relation Extraction with Self-determined Graph Convolutional Network
• Attention Models in Graphs: A Survey
• Entity-Aware Self-Attention and Contextualized GCN for Enhanced Relation Extraction in Long Sentences