A Frustratingly Easy Approach for Entity and Relation Extraction

TL;DR

This paper introduces a simple, effective pipelined approach for entity and relation extraction that outperforms previous joint models on standard benchmarks, emphasizing the importance of distinct contextual representations and global context.

Contribution

The authors propose a straightforward pipelined method for entity and relation extraction that achieves state-of-the-art results and offers an efficient approximation with significant speedup.

Findings

Achieved 1.7%-2.8% improvement in relation F1 on ACE04, ACE05, SciERC.

Validated the importance of learning distinct contextual representations.

Provided an efficient inference approximation with 8-16× speedup.

Abstract

End-to-end relation extraction aims to identify named entities and extract relations between them. Most recent work models these two subtasks jointly, either by casting them in one structured prediction framework, or performing multi-task learning through shared representations. In this work, we present a simple pipelined approach for entity and relation extraction, and establish the new state-of-the-art on standard benchmarks (ACE04, ACE05 and SciERC), obtaining a 1.7%-2.8% absolute improvement in relation F1 over previous joint models with the same pre-trained encoders. Our approach essentially builds on two independent encoders and merely uses the entity model to construct the input for the relation model. Through a series of careful examinations, we validate the importance of learning distinct contextual representations for entities and relations, fusing entity information early in the relation model, and incorporating global context. Finally, we also present an efficient approximation to our approach which requires only one pass of both entity and relation encoders at inference time, achieving an 8-16$\times$ speedup with a slight reduction in accuracy.