Design And Implementation Of Personalized Recommendation System Based On Mobile Internet Friends

With the explosion in the Internet and social networking, it has become increasingly difficult for people to acquire effective information among massive amount of information. Facing the problems of unclear dating demand, inadequate filters of search, a great deal of junk information, how to match users with suitable on-line friends effectively and efficiently is vitally important in social network. According to users' basic information, behaviors and their friends, personalized recommender system combine visible information with invisible information, figure out user's interest, and then guide their dating demand, which will facilitate the searching process considerably and therefore improve users' experience in social networking. Thus, designing a mobile internet-based personalized recommender system is of great significance not only in theory but also in practice.To solve the problems in the friend recommendation module of social network, this article, which is based on collaborative filtering algorithm and content filtering algorithm, integrates user demand and develops a personalized recommender system. This system, first of all, applies training sets of user data to test collaborative filtering and content filtering of different weight ratio, and work out the best one. TF- IDF algorithm is then used to preprocess the basic information of target users, identify the weight of each characteristic item in content-based filtering module. After that, list of recommendation, which is based on content filtering module, could be created through the calculation of users' similarity. Meanwhile, list of recommender system based on collaborative filtering could be worked out according to the relationship between target users. Finally, on the basis of weight ratio of two algorithms, the final comprehensive recommendation list could be figured out by weighting two kinds of recommendation lists. The structure of this system is divided into server and client, using Java development environment and the iOS development platform respectively.