A Privacy Preserving Personalized Group Recommendation Framework

The popularity of Online Social Networking Sites (OSNs) is increasing dynamically. However, an information overload problem has troubled many users as thousands of groups are created each day. Cohesion is one of the widely adopted approach in social networking sites to ensure better accuracy in recommendation systems. Implementation of recommender systems however publishes social network data in one way or another, preserving privacy in publishing social network data becomes an important concern. With some local knowledge about individuals in a social network an adversary can attack easily. Existing works on identity privacy protection on social networks make the assumption that all the user entities are private and ignore the fact that there are considerable amount entities whose identity is public. In this paper we propose a noble approach of group recommendation preserving the identity of user from unauthorized attacker based on the concept of k-anonymity.