基于改进协同过滤算法的个性化美食推荐APP开发

Development of Personalized Food Recommendation APP on Improved Collaborative Filtering Algorithm

为了改善基于个人喜好或营养成分等单一特征的传统协同过滤算法推荐效果不好、缺乏实时性及使用不方便等问题,将中医体质、地方口味、营养成分等多种特征相结合,利用基于Spark Streaming的协同过滤算法,设计一款基于Android平台的智能饮食推荐APP。实验结果表明,改进协同过滤算法能够大幅改善推荐效果,该APP不仅可为用户推荐符合身体健康需要,且合乎口味的饮食,而且能够较好地满足用户对实时性与便捷性的要求。

In order to improve the recommendation effect of traditional collaborative filtering algorithms based on individual preferences or nutritional components,lack of real-time performance and inconvenience in use,a collaborative filtering algorithm based on Spark-Streaming platform is used to realize online real-time recommendation according to the characteristics of traditional diet,traditional Chinese medicine constitution and nutritional components.On this basis,a smart diet recommendation APP based on Android platform is designed.Practice has proved that the improved algorithm based on Android can greatly improve the recommendation efficiency and provide convenience.This APP can not only recommend a healthy and tasty diet for users,but also meet the real-time and convenient requirements of users.