Research And Implementation Of Personalized Food Recommendation System Based On Android

With the development of Internet technology and the improvement of its application level,the amount of data on the Internet has ushered in an explosive growth.The Internet has met people's needs for food,clothing,housing,transportation and so on in various forms,and at the same time has brought about the problem of "information overload".In order to improve user experience and improve the efficiency of information discovery,this problem is generally solved by the following three methods: content classification,search engine and recommendation system.Among the three,the recommendation system is the most widely used.The main function of the recommendation system is to predict the potential preferences of users based on the information of item attribute and user behavior,so as to recommend the items they may like.Collaborative filtering technology is the most commonly used method in personalized recommendation,which provides recommendation results to users through the connection between users and projects.Although collaborative filtering technology has been widely used in e-commerce platforms,it still has some problems such as incomplete perception of users' interests.In this paper,users' preferences are perceived through various factors such as merchants,articles and orders,so as to improve their participation.First of all,the preference score of users to merchants is calculated by the historical behavior records of users and merchants and articles.Secondly,the order matrix is generated by users' orders to merchants.In order to solve the problem of data sparsity which is common in recommendation,it is transformed into order preference matrix and the user similarity is calculated.The preference of users to the category of the order is predicted by the rating of the neighboring users on the category.Finally,several factors were combined to get the recommended results.The experimental results show that the proposed algorithm is more accurate than other two algorithms.This paper based on the Android platform to carry out the design,and the implementation of the recommendation system.First,the requirements of the system are analyzed and the system architecture is obtained.Then,carry on the function design and the database design.Finally,in the system to achieve the merchant,article,order search and display function and user information maintenance function,when the user refreshes the home page,call the recommendation algorithm for users to recommend food.The system has been tested and the test results show that thefunctions are in line with the expectation.