A Brief Survey on De-anonymization Attacks in Online Social Networks

Nowadays, online social network data are increasingly made publicly available to third parties. Several anonymization techniques have been studied and adopted to preserve privacy in the publishing of data. However, recent works have shown that de-anonymization of the released data is not only possible but also practical. In this paper, we present a brief yet systematic review of the existing de-anonymization attacks in online social networks. We unify the models of de-anonymization, centering around the concept of feature matching. We survey the de-anonymization methods in two categories: mapping-based approaches and guessing-based approaches. We discuss three techniques that would potentially improve the surveyed attacks.