Medical Recommendation Model Considering Patient Preference Diversity

The problem of asymmetric information between doctors and patients as well as imbalance between supply and demand growth has always occurred in the medical field.The proposal of Healthy China strategy and Digital China strategy promotes the development of online medical services,which is conducive to solving this problem.However,due to information overload,most patients are limited by their own knowledge level,and still face the challenges of filtering irrelevant information and obtaining information accurately in the process of selecting a doctor.Therefore,intelligent recommendation model has become the research hotspot.The current research on intelligent medical recommendation mainly considers the matching of patients’ disease and doctors’ expertise,while patients’ medical choice decisions are also affected by the perception of doctors’ service performance.Therefore,this paper focus on the intelligent medical recommendation model considering the diversity of patients’ medical choice preferences,and analyze the characteristics of patients’ medical choice preferences with continuous use behavior,such as preferences for service quality,knowledge contribution,online word-of-mouth and treatment experience,etc.Based on this,the key characteristic factors of patients’ decision preference are extracted,and feature variables were constructed based on graded diagnosis and treatment.Furthermore,the original values of feature variables are converted to standard scores,and the comprehensive scores are calculated by entropy-weighted TOPSIS multi-objective evaluation model based on relative status numbers,and the scores are used as target variable.And then,the data relationships between feature variables and target variable are mined using the technique of support vector regression algorithm optimized by the improved sparrow search algorithm.In this way,an intelligent medical recommendation model is constructed to realize the score prediction,and the doctor recommendation set can be generated according to the score prediction.The results of the empirical analysis show that the overall performance of the doctors recommended by the intelligent doctor recommendation model is better and the recommendation accuracy rate has been improved.The results of this paper are conducive to reducing the information search cost and decision-making cost for patients,as well as provide reference for online medical platforms and doctors to improve their service quality,thus promoting the efficiency of medical services and alleviating the conflict between medical supply and demand.