A method for community recommendation for social networks

The increasing use of social networks has been beneficial in our daily lives. In this paper we consider the problem of recommending a user centric community based on a user´s interest topic, location, and relationship network. Our framework can address some significant socio-psychological issues and also allows to understand information propagation. We consider a user´s network structure to calculate the proximity between any two users. We suggest a community recommendation procedure that recommends only the users who are interested in socialization based on several features of a user profile. The recommendation set prunes out bots and cyborgs and it consists of human users only. We have used several properties to analyze a user. We have shown that these properties can be leveraged to improve the performance of the approach in online social networks. We have conducted several experiments using Twitter data. The experimental results illustrate the effectiveness of our approach.