Research On Dietary Intervention Based On Multi-Dimensional Characteristics

Health is the most basic condition for human survival and development.The four cornerstones for maintaining health include “balanced diet”,“appropriate exercise”,“quit smoking and drinking”,and “mental health”.It can be seen that balanced diet occupies a dominant position.At the same time,a balanced diet can maintain good health and prevent chronic diseases,which has become a hot topic of discussion.With the development of social economy and the continuous improvement of the quality of living standards,the user’s diet structure has undergone earth-shaking changes.All the foods are fine rice noodles,and many chronic diseases are actually eaten bit by bit.People eat well is not equal to nutrition,not equal to balanced nutrition,only balanced nutrition can bring health.So how to provide people with a reasonable and balanced diet has become the responsibility of nutritionists all over the world,but because of the limited number of nutritionists and the uneven level of medical practice,it is not convenient for people to understand their eating status portablely.With the development of the mobile Internet,people began to get lost in the vast amount of dietary information and could not find suitable content.Diet recommendations can filter a wealth of information to find content that meets user requirements.Therefore,this paper proposes a reasonable and effective multi-dimensional feature of dietary recommendation methods,which is designed to help people complete the nutritional balance of diet and dietary intervention.The main contents of this paper are as follows:1.Based on the recommendation of the user’s portrait.In this recommendation method,this paper mainly based on different populations,diseases and occupational environment,formulated dietary recommendations for nutrient intake of various attributes,and proved the recommendation by comparing the distribution of nutrient elements in the recommended dishes.The effectiveness of the method and the ability to resolve whether the current user status meets dietary standards.2.Based on the recommendation of TCM constitution.In this recommendation method,depending on the difference in physical fitness of different users,the degree of influence of the body’s internal organs is different.In this paper,the LP algorithm model is used to solve the corresponding mentoring effect of the dish,so it is recommended to adjust the body of the body.The performance index of the algorithm model is 0.0654,while the macro F1 score and micro F1 score are 0.91677 and 0.90637.3.Recommendations based on dietary behavior.Based on the user’s recent dietary record,the clustering algorithm and the improved distance calculation method are used to mine and analyze the user’s nutritional structure and taste preference,and provide dishes that meet the user’s nutritional balance and satisfy the user’s taste.4.The experimental platform was built.This paper builds an experimental platform to implement the dietary recommendation algorithm,and deploys the proposed recommendation algorithm on the experimental platform for the demonstration of the research results.The experimental platform automatically evaluates the physical state and generates the physical report through the questionnaire information submitted by the user.At the same time,the daily eating behavior of the user is collected as the data support of the system,the recommendation algorithm is optimized,and the diet is reasonably recommended.