Research On Data Release And Mining Of Social Network Based On Differential Privacy

With the arrival of big data era as well as the rapid development of information technology,the social network is developing rapidly at an unprecedented scale,and the amount of users' information in social networks is also increasing every day.Massive data spread across every platform of social networks.Social networks contain a large number of personal privacy data,as well as the sensitive information between users and each other,so there exists a lot of security risks to the user information in social networks.In social networks,making friends,information exchange,or sharing resource may bring to the risk of privacy leakage.In social networks,how to protect the sensitive data become particularly important during data release and data mining,which is related to the information security of both the individual user and the whole social network.In view of the traditional method only dealing with the static data but the data in social network growing dynamically,a dynamic clustering algorithm based on differential privacy is proposed,which can dynamically deal with the incremental data in the social network and also can handle the massive information on the basis of differential privacy technology.This method not only improves the efficiency of clustering but also better protect the privacy of sensitive information.For the cross-attack and combinatorial attacks may exist in social networks,a LBS group nearest neighbor query privacy-preserving algorithm based on differential privacy is proposed.A group construction method is proposed,as well as the group privacy budget allocation mechanism and position confusion policy is adopted in the algorithm.And queries of the nearest neighbor group converted to the queries of group centroid which is applied to the whole privacy protection.It is proved that the method can effectively resist the cross attack and combine attack of malicious nodes.Based on the relationship between privacy-preserving and trust,and the different degrees of trust corresponds to the different degrees of privacy-preserving.In order to effectively evaluate the degrees of trust between two different nodes in social networs and combine with the technology of differential privacy-preserving.A privacy-preserving technology which can effectively combine the trust assessment mechanism with the differential privacy protection technology is adopted to better protect the privacy information in social networks.Finally,a brief prototype system is designed on the basis of the theoretical research,which is data publishing and mining system based on differential privacy,it not only can enhance the use of social networking experience,but also can effectively protect the information security of social networks.