Research On Intelligent Guidance System Based On Improved Logistic Model

With the development of computer technology, medical services sites is increasing, online booking gradually replace the traditional way to register. But when patients make an appointment on the Internet are unable to get professional help, it is always exist blindness. At this stage, most medical guide system based on online quiz, but it’s essence is still the traditional artificial guide system. Not only increase labor costs, but also bring unpredictable waiting time. Therefore, how to make use of the information technology and disease diagnosis rules, to set up scientific intelligent guide system. According to the symptoms of the patient’s information can be retrieve possible disease rapidly, guiding patients with accurate registration, become the urgent need of the most hospitals and medical service sites. In this paper, based on the defects of current medical guide system, put forward a kind of special medical guide system based on disease symptoms recommendation and disease similarity matching.In this paper, the design intelligent guide system is mainly including four modules, the patient symptoms of information pretreatment, disease symptoms recommend, symptoms weight of computing, disease reliability calculation. Patient input in the main interface according to their own situation, the patient symptoms of information pretreatment using word segmentation which development by the Chinese academy of sciences, the patients of non-standard participle and synonym match natural language information; In view of the traditional intelligent guide patients input less symptoms, diseases recommendation results poorly. In this paper, proposed a disease recommendation algorithm based on comprehensive symptoms co-occurrence degree, batch asked whether the patients have the disease in the frequency of symptoms at the same time. Whether important or not, in the disease symptoms appear only once, put forward a Logistic algorithm based on comprehensive symptoms co-occurrence degree dimension, can calculate the weight values of symptoms in the disease. Through the above three modules, combining the uncertainty reasoning thought several possible disease, then based on the improved disease similarity calculation model for matching.In order to reflect the proposed methods based on the accuracy of the disease symptoms recommended intelligent guide results and effectiveness, using the same electronic medical record data set respectively in the traditional guide system, and the proposed guide comparison experiment. Experiments show that, the results recommended guide system on the accuracy and stability are improved greatly. Its show that the system on guide results more closer to the doctor’s true judgment, closer to the real demand of patients.