An Improved Approach to the Construction of Chinese Medical Knowledge Graph Based on CTD-BLSTM Model

Abstract

In the process of constructing the knowledge graph, entity recognition and relationship extraction are not only the most fundamental but also the most important tasks, and the effect of their model directly affects the final result of the graph. To establish a more refined knowledge graph, this paper proposes a model of extending Bi-LSTM structural units with Double-word vector and combining Semi-supervised Co-training method. The improved model is used in Chinese named entity recognition and entity-relationship extraction in the Chinese medical field, named Co-Training Double Word embedding conditioned BLSTM (CTD-BLSTM). Experiments show that the CTD-BLSTM model obtains higher accuracy and recall rate than BLSTM in the Chinese medical named entity recognition and entity-relationship extraction. It performs better recognition and adaptability to support the construction of the knowledge graph.