The Diabetic Recommendation System Based On An Improved Collaborative Filtering

Diabetes is a chronic disease, there is no medical means to completely cure it, but diabetes could be controlled by the artificial intervention. Diet control is one of the main means of the treatment of diabetes, its main role is to reduce islet burden of the patients, maintains the blood glucose and lipids at a normal level, coupled with the use of drugs, effectively improve the patient’s health. Traditional diabetes diet is arranged by specific doctors or dietitians, it is hard to manage when the number of patient increases, what’s more, the types of food is few and it’s hard to meet the flavor of patients. Therefore, through the computer and artificial intelligence, for patients with diabetes to recommend a diet is badly in need of solution.In this paper, we researched on collaborative filtering recommendation, to solve the cold start problem in collaborative filtering applied case–based reason, maintain the CBR case library through forgetting mechanism. When the CBR system has been built, the unscored items were deduced by CBR system, and items were recommended with TOP-N used –based collaborative filtering. Experimental results show that the proposed collaborative filtering combined with case-based reasoning could ease cold start problem.Finally, we developed the recommend system for diabetes and doctors by the improved recommend algorithm and android platform. Not only the demand of patients is satisfied, but also the doctors can monitor the diets of patients through the eating habits and information of patients. It is useful to control the illness and reduce the workload of illness dietitians.