Joint Extraction of Entities and Relations Based on a Novel Tagging Scheme

TL;DR

This paper introduces a novel tagging scheme for joint entity and relation extraction, transforming the task into a tagging problem and demonstrating superior end-to-end model performance on a public dataset.

Contribution

A new tagging scheme for joint extraction that simplifies the process and an end-to-end model that outperforms existing methods on benchmark data.

Findings

Tagging-based methods outperform most existing pipelined and joint learning approaches.

The proposed end-to-end model achieves the best results on the public dataset.

The approach simplifies joint extraction by converting it into a tagging problem.

Abstract

Joint extraction of entities and relations is an important task in information extraction. To tackle this problem, we firstly propose a novel tagging scheme that can convert the joint extraction task to a tagging problem. Then, based on our tagging scheme, we study different end-to-end models to extract entities and their relations directly, without identifying entities and relations separately. We conduct experiments on a public dataset produced by distant supervision method and the experimental results show that the tagging based methods are better than most of the existing pipelined and joint learning methods. What's more, the end-to-end model proposed in this paper, achieves the best results on the public dataset.