Research On Surgical Intelligence Question Answering Based On Knowledge Graph

With the advent of the era of intelligence,users have greatly increased demand for medical knowledge retrieval,and are accustomed to human-computer interaction in a simpler and natural way.Therefore,it is necessary to use natural language processing technology to combine user’s intention with knowledge search to help users save time and improve user interaction experience.Aiming at the problem that there are many complex entities in the entity recognition process in the medical field and the feature extraction ability of a single model is insufficient,this paper proposes a Chinese electronic medical record entity recognition model based on deep multi-network collaboration to further improve the accuracy of Chinese electronic medical record named entity recognition,so as to construct an intelligent question answering system based on knowledge graph to help users accurately obtain surgical related knowledge.The main research contents of this paper are summarized as follows:1.Aiming at the serious problem of nested entities in the medical field,this paper introduces the Ro BERTa-wwm-ext pre-trained language model into the task of named entity recognition of Chinese electronic medical records,which can mine deep semantic information in electronic medical records through dynamic masks and full-word masks.At the same time,it used the Global Pointer method to take advantage of the pointer network and processed the nested entities with the idea of global normalization.2.Aiming at the problem of insufficient feature extraction ability of a single model in the medical field,In this paper,four models,Ro BERTa-WWM-Bi LSTM-CRF model,Ro BERTa-WWM-Global Pointer model,Ro BERTa-WWM-LSTM-Global Pointer model and are designed and implemented The Ro BERTa-WWM-Bi LSTM-Global Pointer model is tested on the CCKS2017 and CCKS2019 datasets,and comparative experiments are carried out,aiming to screen out the algorithm model suitable for named entity recognition in medical texts.3.On the basis of the above work,this paper constructs a knowledge graph in the surgical field and builds an intelligent question answering system.Through the system test of the actual use,the results show that this system can effectively deal with the questions proposed by users and has certain practical application value.