A Personalized Privacy Preserving Method for Publishing Social Network Data

Abstract

One of the most important concerns in publishing social network data for social science research and business analysis is to balance between the individual’s privacy protection and data utility. Recently, researchers have developed lots of privacy models and anonymous techniques to prevent re-identifying of relevant information of nodes through structure information of social networks, but most of the existing methods did not cater for the individuals’ personalized privacy requirements and did not take full advantage of distributed characteristics of the social network nodes. Motivated by this, we specify three types of privacy attributes for various individuals and develop a personalized k-degree-l-diversity (PKDLD) anonymity model. Furthermore, we design and implement a graph anonymization algorithm with less distortion to the properties of the original graph. Finally, we conduct experiments on some real-world datasets to evaluate the practical efficiency of our methods, and the experimental results show that our algorithm reduces the anonymous cost efficiently and improves the data utility.