BGPNRE:A BERT-based Global Pointer Network for Named Entity-Relation Joint Extraction Method

• Published in: Ji suan ji ke xue Vol. 50; no. 3; pp. 42 - 48
• Main Authors: Deng, Liang, Qi, Panhu, Liu, Zhenlong, Li, Jingxin, Tang, Jiqiang
• Format: Journal Article
• Language: Chinese
• Published: Chongqing Guojia Kexue Jishu Bu 01.03.2023; Editorial office of Computer Science
• Subjects: Information retrieval; Knowledge representation; named entity-relation joint extraction|bgpnre|global pointer network|bert; Redundancy; Unstructured data
• ISSN: 1002-137X
• DOI: 10.11896/jsjkx.220600239

Named entity-relation joint extraction refers to extracting entity-relation triples from unstructured text.It's an important task for information extraction and knowledge graph construction.This paper proposes a new method-BERT-based global pointer network for named entity-relation joint extraction（BGPNRE）.Firstly, the potential relation prediction module is used to predict the relations contained in the text, filters out the impossible relations, and limits the predicted relation subset for entity recognition.Then a relation-specific global pointer-net is used to obtain the location of all subject and object entities.Finally, a global pointer network correspondence component is designed to align the subject and object position into named entity-relation triples.This method avoids error propagation frompipeline model, and also solves the the redundancy of relation prediction, entity overlapping, and poor generalization of span-based extraction.Extensive experiments show that our model achieves state-of-the-ar