Research And Implementation Of Dietary Recommendation System For Chronic Diseases Based On Data Mining

Proper nutritional intake can not only reduce the risk of major chronic diseases in healthy people,but also delay the disease progression of patients with chronic diseases,reduce complications and improve the quality of life of patients.As different types of chronic diseases need to obey different nutritional treatment guidelines,especially when patients have multiple chronic diseases,the dietary adjustment principles are particularly complicated.In addition,there is a serious imbalance between the supply and demand of medical resources in China,and the number of dietitians and patients with chronic diseases varies greatly.Many patients lack professional nutritional guidance.Therefore,how to efficiently utilize the characteristic information of patients with chronic diseases to carry out scientific and convenient nutritional treatment is the key issue in chronic disease management.Combining with data mining technology and nutrition knowledge,this thesis designs a diet recommendation system based on domain knowledge constraints for patients with chronic diseases.It simulates the process of dietitian making proper dietary plan for patients,and provides targeted and precise nutritional therapy.Since scientific and clear principles of nutritional treatment are the cornerstone of subsequent studies,this thesis proposes a nutritional requirement assessment model based on patient portrait.Firstly,the physiological index data of patients were abstracted to obtain the corresponding labeled portrait.Then,the patient portrait can be used to calculate the daily requirements for core nutrients and to mine individual chronic disease characteristic label.In addition,the data of recipes and ingredients collected from multiple channels has good diversity.At the same time,based on nutrition knowledge,this thesis excavates the characteristic information of each recipe and ingredients and establishes a database of exclusive recipes and ingredients for patients with chronic diseases.In view of the needs of diet management of chronic disease patients,this thesis proposes a comprehensive nutritional treatment scheme combining personalized diet recommendation and diet structure optimization.The former takes into account the patient’s dietary preferences and nutritional dietary rules,and uses the recommendation algorithm to mine the dietary recommendation results that meet the patient’s personalized needs.In this way,patients with chronic diseases have access to multi-scene,precise dietary recommendations.In the latter,multi-objective diet optimization schemes are designed according to the diseases and real-time physiological conditions of people with chronic diseases by comprehensively considering various constraints of protein,energy,potassium and high-quality protein.Compared with dietitians,this thesis can improve the overall efficiency of diet management for patients with chronic diseases and expand the coverage of professional nutrition guidance among patients with chronic diseases.