Research On Large-scale Social Network Privacy Protection Base On K-core

With the increasing number of active users on social network applications,social networks always affect our lives,which makes social network analysis especially important.The publishing of real data sets plays an important role in understanding network structure and analyzing network information flow.Researchers and research institutions explore the potential characteristics of the network by analyzing the information hidden in the social network.However,publishing real social network data sets will bring privacy and security issues,which malicious attackers attack the target users of social networks based on the published data.How to protect the privacy of users and reduce the loss of information in the anonymous social network has become the focus of researchers.In recent years,many research results have been achieved.The number of social network users has grown dramatically,and the traditional stand-alone environment can no longer meet the actual needs of users.A distributed processing graph data framework based on the Pregel model was designed.The algorithm was demonstrated execution efficiency and data availability based on many real social network data sets.The results show that the distributed social network privacy protection method based on k-core can protect the data availability while computing graph data in a distributed manner.First,in order to solve the problem of sensitive edge protection in social network graphs,a random perturbation method for social networks based on k-core is proposed.The algorithm processes the large-scale social network graph based on the Pregel model,and find high-core neighbor nodes to replace the edges.The target is that the core number and community structure does not change after anonymity.The probability of an attacker correctly identifying the connecting edges limits through random perturbations.Then,in order to solve the problem of graph structure destruction by random perturbation algorithm,a random perturbation method based on k-core for sensitive areas is proposed.The Firefly algorithm is applied to social networks.High-influence nodes are searched in parallel based on the Pregel model.The low-influence nodes are gathered around the high-influence nodes to for sensitive areas.The k-core,degree,and PageRank algorithms are used as the firefly algorithm node initialization influence.The Firefly algorithm is used to obtain nodes with different influences.Finally,according to different edge retention probabilities,random perturbations protect privacy and security in sensitive areas.Finally,in view of the existing anonymous algorithms ignoring the influence of nodes after anonymity,a social network node splitting anonymity method based on k-core is proposed.The social network graph is decomposed to obtain a k-core graph.The splitting the anonymous social network of nodes is computed in parallel based on the Pregel model while ensuring that the original node's core number and the influence of the node is unchanged.In order to further improve the strength of the algorithm anonymous protection and Stable community structure,a social network anonymous algorithm for protecting the influence of nodes in the community is proposed according to the community structure.