Research And Implement Of Intelligent Medical Diagnostic System

In recent years, with the rapid development of the Internet, access information from Internet has become a popular way in everyone’s life. In the medical field, people also tend to search on internet when they get sick, so as to understand their symptom and possible disease. However, the information in the Internet is very huge and complex, while the general search engines often do not have the professional knowledge in the medical field, so they can only do shallow keyword search which make it difficult to meet the medical needs of people.This word aimed to establish an aotumated medical diagnosis system based on inferred many times, in order to better help sick people to self-diagnosis. In this system, users descripe their symptoms to system at first, then the system will infer users possible diseases. When system determines that the user information is not enough to diagnosis, the system will ask the user some questions, so as to make more accurate inferences based on user feedback. Finally, the system will provide an analysis report to user about the possible diseases, so as to help user see the doctor.This paper first describles the method and results of the knowledge based which is used in diagnosis system, including crawl medical related web data, using different method based on different type of data to recognize the named entity, such as disease and symptoms, and quantify entities and relationships between entities, and so on. Then this paper describes how to combine word2 vec and semantic analysis to extract symptoms from user’s descripe and feedback. When the system determines that the user feedback is not sufficient to diagnose, the system will calculate the information gain of each symtpom if system ask the user about it, so as to ask the symptom which could provide most information gain. In this process, system will infer the possible disease of user by Bayesian. Finally, this paper test the method of this paper on the crawled data, and introduced the optimization algorithms in the implementation.