A Privacy Analysis Method to Anonymous Graph Based on Bayes Rule in Social Networks

Abstract

With the widespread popularity of social network platforms, privacy leakage has become a focus when users share personal information. As a result, there are so many different privacy preserving methods. However, one of the methods is to add noise to get a published graph, which cannot achieve the privacy preserving of social network completely. In this paper, we proposed a Bayesian privacy analysis model to identify nodes in a published graph. Firstly, we built a general privacy analysis model to explain the main idea of privacy analysis. Secondly, under this model, a privacy analysis method based on Bayes Rule is designed. Finally, experimental evaluation results showed that our method could identify one node of published graphs with some probability. Therefore, our model also provides guidance for designing better privacy preserving methods of social network.