A Hybrid Algorithm for Privacy Preserving Social Network Publication

With the rapid growth of social networks, privacy issues have been raised for publishing data to third parties. Simply removing the identifying attributes before publishing the social network data is considered to be an ill-advised practice, because the structural characteristic may reveal the users privacy. We discuss the current techniques for publishing social network data and define a privacy preserving social network data publishing model with confidence p. Then we devise a hybrid privacy preserving algorithm satisfying the defined model for publishing social network data. Combining the features of k-anonymity with randomization, the algorithm uses the k-anonymous concept to hide the sensitive information into the natural groups of social network data and employs random approach to process the residual data. We conduct the algorithm on several real-world datasets, the experimental results show that our algorithm is practical and efficient. Compared with the related k-anonymity and random methods, our algorithm is stable and modifies the original data less than the existing algorithms.