Defending Against User Identity Linkage Attack Across Multiple Online Social Networks

We study the first countermeasure against user identity linkage attack across multiple online social networks (OSNs). Our goal is to keep as much as user's information in public and meanwhile prevent their identities from being linked on different OSNs via k-anonymity. We develop a novel greedy algorithm, incorporating an efficient manner to compute the greedy function, and validate it in terms of both solution quality and robustness using real-world datasets.