Protecting Sensitive Relationships against Inference Attacks in Social Networks

Abstract

The increasing popularity of social networks in various application domains has raised privacy concerns for the individuals involved. One popular privacy attack is identifying sensitive relationships between individuals. Simply removing all sensitive relationships before releasing the data is insufficient. It is easy for adversaries to reveal sensitive relationships by performing link inferences. Unfortunately, most of previous studies cannot protect privacy against link inference attacks. In this work, we identify two types of link inference attacks, namely, one-step link inference attacks and cascaded link inference attacks. We develop a general framework for preventing link inference attacks, which adopts a novel lineage tracing mechanism to efficiently cut off the inference paths of sensitive relationships. We also propose algorithms for preventing one-step link inference attacks and cascaded link inference attacks meanwhile retaining the data utility. Extensive experiments on real datasets show the satisfactory performance of our methods in terms of privacy protection, efficiency and practical utilities.

The work is partially supported by the National Natural Science Foundation of China(Nos. 61173031, 60973018) and the Fundamental Research Funds for the Central Universities(Nos. N090504004, N100604013).