Research On K-In&Out-Degree Anonymity Of Large-scale Social Networks For Protecting Community Structure

With the rising popularity of the Internet penetration,the development of various social Apps has become increasingly mature,and the number of online users has continued to increase.As of February 2019,the penetration rate of WeChat installations reached 85.8%,the monthly active users reached nearly 1 billion,and the average daily active users reached 648 million.Researchers analyze the community structure of large-scale social network directed graphs,which has important research significance in similar group discovery,group behavior pattern discovery and so on.The actual social network directed graph often involves the user's personal privacy information.Attackers can easily identify the target users of social network through the background knowledge,which leads to the leakage of personal information.First,existing social network privacy protection technologies have low performance in dealing with large-scale social network directed graph data,and anonymous data publishing does not meet the needs of community structure analysis.A large-scale social network KIn&Out-Degree anonymity method is proposed.The method divides the community based on the hierarchical community structure algorithm.The greedy algorithm is used to group and anonymize the k-in&out-degree sequences,and the virtual nodes are added in parallel to achieve k-in&out-degree anonymity.Based on GraphX,the information between nodes is transferred,and the virtual node pair is merged and deleted according to the change of the hierarchical community entropy of directed graph to reduce the information loss.Then,according to the different needs of users in the social network,the proposed K-In&OutDegree anonymity is extended,and a large-scale personalized K-In&Out-Degree anonymity method for the social network is proposed.Set the user's requirements to Lv0~Lv3 privacy protection level,and the large-scale social network directed graph is personalized k-in&outdegree anonymity.Finally,a large-scale dynamic social network K-In&Out-Degree anonymity method to protect the community structure is proposed for large-scale dynamic social network directed graphs.Anonymize the dynamic k-in&out-degree sequence according to the dynamic grouping anonymity rule,and add virtual nodes to construct an anonymous graph in parallel.Based on GraphX,the information between nodes is transferred,and the virtual node pair is merged and deleted according to the change of the directed graph module to reduce the information loss.In this paper,three methods are tested and analyzed by using real social network data sets.The experimental results show that the methods improve the efficiency of processing directed graph data of large-scale social networks and ensure the high availability of community structure analysis when data is published.The methods also meet the needs of different users for privacy protection,and achieve the privacy protection of dynamic social network directed graph data.