Research On The Construction And Application Of Medical Knowledge Garph

Since the reform and opening up,China’s economy has developed rapidly,and Internet technology and computer technology have also emerged and developed,The network is related to people’s lives or work.In the era of explosive growth of Internet information,traditional information retrieval methods have been unable to meet the daily needs of users.How to accurately and quickly search for the information required by users in a large amount of complex information has become a hot issue in the field of information retrieval.The concept of knowledge graph proposes a new solution for this research field.It can process massive amounts of data,then extract the entities contained and the relationships between them and store them in a database in an intuitive way.In this article,the current popular entity recognition technology-BiLSTM-CRF is applied,which aims to extract the names of drugs,disease names and symptom names in the medical field text.CRF is a classic word segmentation method in the field of machine learning that combines it with bidirectional LSTM technology in deep learning.After the CRF is connected to the bidirectional LSTM model,the global optimal output sequence generated by the LSTM model is obtained and reused to better fit the data.Experiments show that the combination of the two has achieved satisfactory results in terms of effectiveness and efficiency.After that,the current popular question-answering system is used as the prototype,which aims to apply the data on health care that can be collected.Firstly,the Neo4j graph database is used to store data,and then the HanLP and Naive Bayes classification technology is used to implement word segmentation and matching of user problems.Finally,the results retrieved in the database are visually presented through the question-answering system.The experimental results show that the named entity recognition technology based on BiLSTM-CRF has certain efficiency in dealing with medical field texts.At the same time,the classifier based on the naive Bayesian algorithm has a high accuracy in question matching,which proves the usability of the question and answer system.The realization of this system can provide auxiliary medical services for doctors,so it has certain practical significance.