Research On Social Network Privacy Protection Based On Markov Cluster Algorithm

Social network contains the interaction between social members,which constitutes the structure and attribute of social network.Social network is a new application mode under the Internet background.The distribution of social network data has dynamic characteristics.The interaction of social network contains a large number of personal privacy information.In the release process of analysis mining,the user's personal privacy information will be leaked.How to carry on the effective privacy protection to the social network data publication has the far-reaching theory significance and the practical value in the field of information security.This dissertation focuses on the research of privacy protection in social network data publishing,and the main works are listed as follows:(1)Aiming at the privacy protection in the data release of static social network,a social network differential privacy data publishing method based on MCL(MDPA)is designed.MDPA satisfies ?-differential privacy.First,a differential privacy protection model is constructed to adapt to the social network graph features.Secondly,the Markov clustering algorithm is used to cluster the social network graphs,and then generate the triple function of edge weight information.Finally,we generate the privacy-protected social network graph by injecting the noise obeying Laplace distribution into the weight of the edge of the cluster node.(2)Aiming at the problem of privacy protection in data publishing in dynamic social network,this dissertation improve the privacy protection algorithm in static social network data publishing and design a dynamic social network data publishing algorithm(DDPA)that satisfies the ?-differential privacy.First,the edge weight information that changes with the dynamic change of the social network graph are identified,which include adding edges and modifying edges.Then generate the triple function.Secondly,we add the privacy protection budget that satisfies ? and build ?-weight vector.Finally,generate the privacy-protected social network graph.(3)In view of the above two aspects,the simulation experiments are carried out.Theexperimental results show that the MDPA algorithm satisfies the user's differential privacy requirements in the social network and improves the data utility.The DDPA algorithm adapts the dynamic characteristics of the social network,greatly greatly improves the efficiency of the algorithm,and ensures the reduction of the loss rate of weight information.