A Research On Intelligent Voice Question Answering Technology For Seeking Doctors And Medical Advice

With the development of the Internet and the improvement of people’s living standards,the contradiction between the various services in my country’s existing medical field and the growing needs of the people has become more and more prominent,smart medical treatment can effectively improve various services.Therefore,the development of smart medical care is of great significance to alleviate the above-mentioned contradictions.Professional websites in the medical field have accumulated a large amount of medical data during the development process.Therefore,the use of knowledge graph technology to extract effective information from the data and apply it has become one of the keys to the development of smart medical care.The application of intelligent question answering system based on knowledge graph in the professional field can more accurately understand the user’s intention and return the question result.In this paper,considering factors such as data size and feasibility,a category of "liver disease" that is currently less studied in the medical field is used as the research object to construct a knowledge map based on liver disease,and on this basis,develops an intelligent voice question answering system about liver disease which provide a tool to assist in the treatment of liver disease and a channel for obtaining professional knowledge to patients.The main research work of this paper is as follows:(1)Improve BERT model to improve the accuracy of the named entity recognition model.Compared with the performance of the current mainstream named entity recognition model applied to the actual annotated data set,the Bi LSTM-CRF model with better performance is selected to apply to the liver disease entity recognition in this paper.The word embedding algorithm is switched to a BERT model that has better application effects on small data sets and can distinguish the meaning of the same character in different contexts.Improve the accuracy of the application of the entire model on specific data by improving the random coverage method of the BERT method.(2)Design the schema structure.By analyzing and summarizing data,the schema structure used in this article is designed,the knowledge graph and intelligent question answering system are constructed based on this.The schema structure can be applied to other similar medical fields.(3)Construct a knowledge map of liver disease.Due to the lack of research on liver disease knowledge maps,there are fewer data sets in related fields.Therefore use crawler technology to crawl liver disease related data on professional medical websites,perform entity recognition knowledge extraction and entity alignment knowledge fusion operations on the data,finally store the data in the graph database Neo4 j and build a relatively complete liver disease knowledge map.(4)Design and implement an intelligent voice question answering system.First,the user’s voice is converted into text,and certain word segmentation preprocessing is performed on the input text;then,the syntax is analyzed,the keywords in the question sentence are extracted and the type of the question sentence is judged,thereby converting the question sentence into three Tuple structure;finally match the triples with the Cypher query template,query the results in the knowledge graph and return them to the user in the form of text or voice.