Research of Chinese relation extraction based on BERT

Le résumé fourni par la source

Facing massive text data on the Internet, it is meaningful to extract valuable information for building knowledge graphs and knowledge engines. Relation extraction is a widespread issue in current research. This paper proposes the BERT-BiLSTM-CRF model for Chinese relation extraction. This paper focuses on ten types of Chinese named entity relations, and the model achieves excellent results in the extraction task of these ten relations. This paper first constructs word vectors using the pre-trained BERT encoding model and then extracts features using a bidirectional long short-term memory network (BiLSTM) followed by semantic encoding processing. Finally, this paper uses a conditional random field (CRF) to predict relation categories and complete Chinese relation extraction. Experiments show that the performance is improved after applying the model.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.