Research On Privacy Protection Of Social Network

In recent years，various social networking products are introduced into the Internet, from Happynets, Renren to popular MicroBlog. Social network is changing our lives. Large amount of personalinformation is accessible on the network, so that protecting the privacy information, especially thesocial relationship data, is becoming increasingly important.This thesis aims to do research on the privacy protection of social network from the aspect ofdata dissemination. In this thesis, at first, the latest research status of the privacy protection of socialnetwork is summarized, the basic conception and trait of social network are introduced, the attacksare analyzed, and anonymous dissemination methods are summed up. Then, by improving existingmethod, two new anonymous methods--k-symmetry anonymity method and cube mappingsymmetry anonymity method are designed. K-symmetry anonymity method aims to resist the attackof structure knowledge. It respectively duplicates each node, so that each set which belongs to theresult of equivalence class has k nodes. Through duplicating nodes in such way, this method can letthe probability of distinguishing one node not more than1/k. Cube mapping symmetry anonymitymethod aims to resist the background knowledge attack. This anonymous method can realize theprivacy protection because nodes and lines of the social network are mapped in the cubes. Theavailability and privacy degree of the two anonymous methods have been analyzed throughemulating and instance. Besides, a recovery algorithm which can recover the original socialnetwork graph and is used in analyzing availability of k-symmetry anonymity is designed. In theend, this thesis implements the two anonymous methods by the data of e-mail’s social network.Besides, it evaluates their usability and verifies the availability.This thesis has done beneficial work in the privacy protection and data anonymitydissemination of the social network.