Identification of Network Communities and Assessment of Privacy Using Hybrid Algorithm

Social networking platforms have rapidly transformed the method of interaction among billions of users since the last decade. Such user generated content plays a significant role in decision making strategies. It is crucial to preserve concealment of such personal data prior to assimilation with third party developers. Conventional security measures adopted by social networking sites often fail to preserve data from adversary threats. Towards this objective, the paper proposes a hybrid algorithm based on randomization and k-anonymity, which pursues to preserve privacy of users in network communities by decreasing the probability of information loss. The key idea behind the algorithm is to randomize vulnerable attributes within communities by preserving their degree followed by anonymizing the randomized values. The evaluation of hybrid algorithm based on association rules shows that this method improves data integrity compared to other methods.