Joint extraction of entities and overlapping relations using source-target entity labeling

作者：

Highlights：

• Joint extraction model of entities and overlapping relations is proposed.

• Source-target BERT, based on pre-trained model, is used to assign entity labels.

• OREM, a three-step model to extract the unlimited number of overlapping relations.

• Proposed model on two public datasets exhibits greater effectiveness.

摘要

•Joint extraction model of entities and overlapping relations is proposed.•Source-target BERT, based on pre-trained model, is used to assign entity labels.•OREM, a three-step model to extract the unlimited number of overlapping relations.•Proposed model on two public datasets exhibits greater effectiveness.

论文关键词：BERT,Entity recognition,Overlapping relation extraction,Joint extraction model

论文评审过程：Received 10 August 2020, Revised 9 January 2021, Accepted 3 March 2021, Available online 13 March 2021, Version of Record 30 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114853