The Research Of Privacy Protection Technology For Users In Social Networks

With the gradual popularity of the social networks,more and more social network data needs to be published for scientific and commercial purposes,which will undoubtedly lead to the leakage of users'privacy information in social networks.At present,many researches at home and abroad focus on the technology of preserving privacy information of users in social networks,including the existence of users,the sensitive labels of users,the relationship between users,the weight of relationships,the kinds of relationships,the structure of social networks and other technology.For different background knowledge of attackers,attackers can use these background knowledge to launch different types of attacks.In recent years,there are a lot of researches focusing on different attack models in social networks.And for a specific attack model,the corresponding algorithm is proposed to protect the users'privacy information.This paper focuses on a new type of privacy attack model——non-sensitive label pair attack.In the non-sensitive label pair attack,the attacker can re-identify a pair of friends by using the non-sensitive labels of two vertices connected by an edge.We present a new anonymity algorithm,called Non-sensitive Label Pair k~2-anonymity(NLP k~2-anonymity)which ensures that there exist at least k-1 other vertices such that each of the k-1 vertices also has an incident edge of the same label pair and reduces the probability of a vertex being re-identified to less than or equal to 1/k.After reaching the results of Non-sensitive Label Pair k~2-anonymity,if the vertices with the same non-sensitive label pair connect to the vertices of the same non-sensitive label that have the same sensitive label,the sensitive label information of the vertices will be leaked.Therefore,on the base of Non-sensitive Label Pair k~2-anonymity,we fuse into l-diversity algorithm and further anonymize the social network.We call this anonymity algorithm which we propose non-sensitive label pair k~2-anonymity sensitive label l-diversity(NLP k~2-anonymity SL l-diversity)algorithm.Finally,it demonstrates that the anonymity algorithms that we propose can preserve the users'identity,the users'privacy information and utility of social networks effectively through the results of the experiments.