Research On Social Network Data Publishing Based On Combined Differential Privacy

With the development of the Internet,there are more and more users of the Internet,and the amount of data is getting larger and larger,forming the social network of the Internet world in the Internet.Social networks contain information about users on the Internet and interaction information between users.Social networks have static and dynamic characteristics.Social networks contain a large amount of user information,including the privacy information of many users.During the process of data sharing and data analysis,there is a risk of privacy leakage.Therefore,in the current Internet era,how to effectively protect the privacy information released by social network data is a hot issue in the field of information security.Aiming at the problem of privacy protection of social networks,the main work is as follows:(1)Aiming at the unbalanced privacy protection of data in the static social network,a combined differential privacy data publishing method in the static social network is proposed.Using Markov clustering(Markov cluster algorithm,MCL)and Chameleon hybrid clustering social network graph is divided into several clusters,using differential privacy parallelism of the combination of the characteristics of distribution of individual privacy for each cluster budgetε,structure meet the(maxε_i)-difference weight vector of privacy model,to all the clusters in the edge to add Laplace noise,get data privacy protection after static social network diagram.(2)Aiming at the problem of slow iteration of privacy protection methods in dynamic social networks,based on the use of B+trees,a combined differential privacy data publishing method in dynamic social networks is proposed.Use the B+tree to index the edges of the social network graph,divide the index data of the B+tree,and assign differentεvalues and add Laplacian noise to the data according to the characteristics of the differential privacy parallelism combination to achieve the overall efficiency after data privacy protection.Usability and local strong protection;the index of the B+tree is used to quickly locate the updated information during iteration,and differential privacy protection is achieved in the rapid iteration of the dynamic social network.(3)Simulation tests of the two methods.The experimental results show that the privacy protection method in the static social network effectively balances the protection intensity of the data and improves the utility of the data.In order to improve the execution efficiency of the algorithm,it can better adapt to the dynamic nature of the social network,and ensure that the information loss rate is maintained at a low level,which improves the data utility of the dynamic social network.