Research On The Construction Technology Of Disease Knowledge Map For Orthopedics Consultation Of Traditional Chinese Medicine

The popularity of Internet and artificial intelligence in various fields and people's awareness of health care are increasing day by day,and the demand for medical information in the whole society is also increasing."Intelligent medicine" will be the development trend of the future medical industry.It can not only provide mass medical information for the public,but also solve the problems of uneven medical resources and complicated medical procedures.Knowledge graph in medical field is the cornerstone of intelligent medicine,which makes the exchange of medical information resources more convenient and provides data support for efficient and convenient medical services.By collecting and processing electronic medical records of orthopaedics in Kunming Hospital of Traditional Chinese Medicine and related knowledge from the Internet,this paper studies the construction of knowledge graph of orthopaedics inquiry platform of traditional Chinese medicine.Firstly,extracting entity relationship and obtaining the "entity-relationshipentity" triple which is necessary to construct knowledge graph as the basic data.Aiming at the characteristics of electronic medical records of traditional Chinese medicine,a method of joint extraction of entity relations based on entity relations labeling strategy is proposed.According to the labeling strategy formulated under the guidance of relevant experts,the electronic medical records are processed and labeled first,and the in-depth learning model based on Bi-LSTM(bidirectional long-term and short-term memory network)is used for training and learning to extract entity relations more accurately,which overcomes the traditional methods.Problems such as large errors.After obtaining the entity relationship,the entity relationship is perfected and supplemented through the Internet acquisition and comparison.The experimental results show that the method has a higher recall rate and F1 value.Secondly,in view of the acquired entity relationship,there exists the phenomenon of duplication of "multi-word" meaning,that is,multiple disease names refer to the same disease,which requires co-referential resolution to integrate these entity relations.This paper proposes a knowledge fusion method based on CNN(Convolutional Neural Network Model).The model is used to calculate the similarity of symptoms corresponding to two diseases,and to judge whether they are duplicated for knowledge fusion.This method solves the problem of sparse medical short text features and difficult to distinguish whether the semantic information is similar or not.The experiment achieves better accuracy and recall rate.In this paper,the research on the construction of knowledge graph of orthopaedics of traditional Chinese medicine is focused on the extraction of entity relationship of electronic medical records of traditional Chinese medicine and the research and exploration of the knowledge fusion of co-referential resolution of "multi-word" phenomenon,and the related design of the construction of orthopaedics inquiry platform of traditional Chinese medicine,as well as the realization of the abovementioned methods of entity relationship extraction and knowledge fusion.