UniRE: A Unified Label Space for Entity Relation Extraction

TL;DR

This paper introduces UniRE, a unified approach to entity relation extraction that simplifies label spaces, improves efficiency, and achieves competitive accuracy across multiple benchmarks.

Contribution

We propose a unified label space and classifier for entity and relation extraction, enabling better information interaction and faster inference.

Findings

Achieves competitive accuracy with fewer parameters.

Faster inference compared to existing models.

Effective approximation decoding method.

Abstract

Many joint entity relation extraction models setup two separated label spaces for the two sub-tasks (i.e., entity detection and relation classification). We argue that this setting may hinder the information interaction between entities and relations. In this work, we propose to eliminate the different treatment on the two sub-tasks' label spaces. The input of our model is a table containing all word pairs from a sentence. Entities and relations are represented by squares and rectangles in the table. We apply a unified classifier to predict each cell's label, which unifies the learning of two sub-tasks. For testing, an effective (yet fast) approximate decoder is proposed for finding squares and rectangles from tables. Experiments on three benchmarks (ACE04, ACE05, SciERC) show that, using only half the number of parameters, our model achieves competitive accuracy with the best extractor, and is faster.