A clustering approach for privacy-preserving in social networks

Abstract

Social networks, in which huge numbers of people spread massive information, are developing quite rapidly. Here people can obtain interesting information much more quickly and conveniently. However, people’s privacies leak easily here too. A lot of works have been done to deal with this problem. Most of them focused on either attribute information or structure information. It is insufficient, because both attributes and structures, including sensitive attributes, are important in social networks, and we need to protect both of them. In this paper, we introduce a novel approach for privacy-preserving considering both attribute and structure information. In particular, sensitive attributes are considered to resist re-identification attacks. Moreover, we define the entropy to measure capability of preserving sensitive attributes.