Sentiment analysis and user similarity for social recommender system: An experimental study

Abstract

Social recommender system has become an emerging research topic due to the prevalence of online social networking services during the past few years. A social recommender model can be considered the combination of a recommender model and a social information model. Many approaches have been proposed to exploit the social interaction or connections among users to overcome the defect of traditional recommender systems assuming that all the users are independent and identically distributed. In this paper, we propose a social recommender system using memory based collaborative filtering models with user-oriented methods as basic models, in which we conduct an analysis on the correlations between social relations and user interest similarities. We also combine techniques of sentiment analysis to get dataset of users with their favorite products; this dataset is the input for the social recommender system. Our experiments on giving recommendations for Facebook users about mobile phones show the efficiency of the proposed approach.