Protecting Sensitive Labels in Weighted Social Networks

With the popularity of social networks, data privacy preserving in social networks has become a hot issue among scholars. An attacker can use a variety of background knowledge to attack against privacy. Most of the present technology on anonymity weighted social network graphs can only deal with edge weight, but cannot be applied to sensitive labels. We consider a new generalization approach for sensitive labels, which can afford utility without compromising privacy. In this paper, we investigate the sensitive label privacy disclosure problem in weighted graph, propose k-histogram-inverse-l-diversity (KH-inv-LD for short) anonymity to protect sensitive label information, and develop a label anonymous approach to achieve this model. Extensive experiments on real data sets show that the algorithm performs well in terms of sensitive label privacy protection in weighted graph.