Personalized recommender system based on social relations

Advent of the Social Web and the ever increasing popularity of Web 2.0 applications, has led to a massive amount of information. Therefore, users have difficulties in finding their desired information according to their interests and preferences. To address this issue, recommender systems have been emerged. These systems try to provide users with the most relevant and suitable information they need by investigating their preferences as well as their demographic information. With the growing development of social networks and the number of users in them, the value of information in these systems has also increased. This information in social networks can be used to improve the precision of recommender systems. In this paper we present a novel recommender system that makes use of user's social relationships in two levels: computing the similarity between them and identifying user's neighbors set. Our experimental results show that the proposed model outperforms Collaborative Filtering (CF) based recommender system in terms of recommendation accuracy.