Research On The Intelligent Medical Guide System Based On Multi-words TF-IDF Algorithm

Currently, due to the growing number of consulting patients, patients have less effective communication time with the doctors from the department of human-aided medical guide, thus, patients were difficult to get detailed answers and human-aided medical guide has been unable to meet the currently demand of medical guide. Most Internet medical service website and mobile App, use on-line human-aided medical guide service actually. Patients have to get a period of unpredictable waiting time for the human-aided medical guide results after they input their symptoms. Therefore,how to build an intelligent medical guide system by computer technology and medical knowledge, this system applies the symptoms of user input to calculate the possibility of disease for users and gives the users reasonable guidance, instead of human-aided medical guide and online human-aided medical guide, has become a strong demand in many hospitals, Internet medical services and Apps. According to the current problems of artificial human-aided medical guide and online human-aided medical guide, this paper proposes an intelligent medical guide system which combination the reasoning thinking of medical diagnosis expert system and the thinking of disease similarity calculation.This paper designed an intelligent medical guide system which consists three subsystems: 1) UI subsystem, this subsystem is realized in the Android operating system, patients are able to input symptoms at anytime and anywhere, and the system can guide the patients. 2) NLP(Nature language symptom) subsystem, this subsystem adopting an open source participle IKAnalyzer and extension dictionary of word segmentation to segments the natural language of symptoms which input by patients,then using literal similarity algorithm of the back center of gravity according to the characteristics of the medical professional vocabulary. 3) MGC(Medical guide calculation) subsystem, this subsystem using the reasoning thinking of medical diagnosis expert system to calculate the initial disease, then convert the results to vector by VSM(Vector Space Model) and calculate the reliability of the diseasesbased on the similarity of the vectors. Finally display the calculation results in the UI subsystem friendly. Because of most patients input more symptoms, this paper proposes the multi-words TF-IDF algorithm to calculates the weight.For the influence of the accuracy and reliability of the intelligent medical guide system by the multi-words TF-IDF algorithm, we achieved the system respectively by the TF-IDF algorithm and the improved TF-IDF algorithm and using the same data set to verity the correctness and reliability of two systems. The related experiments proved that the improved TF-IDF algorithm in this paper increased the correctness and reliability of medical guidance. This proved that the results of the improved TF-IDF algorithm is better than the original algorithm and more close to the correct results.