Relation Extraction with Self-determined Graph Convolutional Network

TL;DR

This paper introduces SGCN, a novel relation extraction model that automatically determines graph structures using self-attention, eliminating reliance on linguistic tools and achieving state-of-the-art results on TACRED.

Contribution

The paper presents a self-attention based method for graph construction in relation extraction, enabling end-to-end learning without linguistic tools.

Findings

SGCN outperforms traditional GCN models on TACRED.

Self-determined graphs improve relation extraction accuracy.

Eliminates dependency on linguistic tools for graph construction.

Abstract

Relation Extraction is a way of obtaining the semantic relationship between entities in text. The state-of-the-art methods use linguistic tools to build a graph for the text in which the entities appear and then a Graph Convolutional Network (GCN) is employed to encode the pre-built graphs. Although their performance is promising, the reliance on linguistic tools results in a non end-to-end process. In this work, we propose a novel model, the Self-determined Graph Convolutional Network (SGCN), which determines a weighted graph using a self-attention mechanism, rather using any linguistic tool. Then, the self-determined graph is encoded using a GCN. We test our model on the TACRED dataset and achieve the state-of-the-art result. Our experiments show that SGCN outperforms the traditional GCN, which uses dependency parsing tools to build the graph.