Deep learning Based Patient-Friendly Clinical Expert Recommendation Framework

In recent years, with the popularization of the Internet and the development of technologies such as big data analysis, people's demand for mobile medical services has become more and more urgent, which is manifested in determining their diseases based on symptoms and selecting hospitals with better service quality according to the illnesses and doctors. An inquiry recommendation system is designed and implemented based on knowledge graphs and deep learning technology to solve the above problems. Based on the open medical data on the Internet, a “disease-symptom” knowledge map is constructed to help users self-examine according to symptoms. The knowledge map embedding model trains the embedded vector representation of entities in the knowledge map. The most similar is selected according to the Euclidean distance similarity of the vector. The disease entity enriches recommendation options, and the two are combined to achieve disease diagnosis services. At the same time, based on social media comment data, combined with the existing medical service quality evaluation indicators, the deep learning analysis method is used to automatically give a multi-dimensional score of the doctor's service quality and provide users with the doctor and hospital recommendation services. Finally, by constructing test sets and designing questionnaires, it is verified that the accuracy rates of disease diagnosis service and doctor-hospital recommendation service are 74.00% and 90.91 %, respectively.