Research On The Intelligent Medical Guide System Basing On The VSM Weight Improved Algorithm

With the development and popularization of the Internet, there is more and moremedical services website witch function is more intelligent, online registrationfunction has also been widely applied, so that the user can not out of the house canmake an appointment online, convenient for users to visit. The traditional medicalguide mainly uses artificial methods, the patients go to the hospital for medicaltreatment are often unable to determine to registration which doctor and department,they need to through describe their symptoms to ask the nurse which doctordepartment should be registered, the user through the online registration can’t bedirectly ask professionals to get help, there is blindness, therefore, how to through thecomputer system search queries possible diseases, and guide users registeredcorrectly to instead of manual medical guide has become the urgent demand for manymedical websites. In this paper, according to the current research status of intelligentguide medicine and based on the combination of the medical diagnosis expert systemand network technology, the intelligent guiding system combining with the intelligentexpert system inference and the VSM (vector space model) was proposed.This paper first analyzes the problem which system implementation need tosolve, mainly include two parts are natural language processing and diseasesinference and similarity calculation. The processing of natural language primarilyuses the existing of Pangu word segmentation to achieve, and then uses the center ofgravity moving after to the literal similarity algorithm to identify synonyms. Thispartly solved the problem of users professional vocabulary is limited, and makeshuman-computer interaction more convenient. This paper uses uncertain inferencethought of expert system and vector space model for the disease preliminaryreasoning and similarity calculation, in the process of the system implementation thatall weights of symptoms by the TF-IDF algorithm calculate are identical, in order tosolve this problem this paper uses the user attention values of Baidu index as thecommon degrees, at the same time, according to the characteristics of the disease diagnosis to improve the TF-IDF algorithm, then uses this algorithm to calculate theweights of symptoms.This paper uses the improved algorithm and the former algorithm to separatelyimplementation system, and then enter the same content in the two systems tocomparative experiments to verify the accuracy of the improved algorithmimplementation system is higher, and the realize effect is better.