Design Context Aware Activity Recommender System in Social Networks Using Clustering Technique

Nowadays, social networks are widely used by everyone. So, it is necessary to do appropriate and situation aware activities in these networks to gain benefits. In this research, a context aware recommender system has modeled for using in social networks. This system makes its recommendations for user based on behavior and activities of her friends in the same situation in social network. In other word, this modeled recommender system uses collaborative filtering algorithm. All the connections of user in social network, containing direct and indirect, are considered for recommending by recommender system; but, based on connection type and its distance to user, proportional factor is assigned. For example, the factor of friends of user is 1, and the factor of friends of friends is ½. At last, by implementing this model on an appropriate data set using data-mining and neural networks techniques and then, calculating its precision, the usefulness of its recommendations are evaluated.