Dynamic Trust Model For Social Networks

With the development of Internet and increasing demand of communication among people, socialnetworks have been playing an important role in our lives. As a huge country with more than one billionpeople, we also witnessed the fast growth of its own social networks. Social networks based on Web2.0have gained popularity among people thus changing our daily lives. Almost every online social networkprovides open API interfaces and, for this reason, sharing of information and search engines could bepossible among these huge distributed networks. However, trust has become a bottleneck to block thefurther development and therefore more and more people pay their attention to this issue. It has an obvioussignificance on the development of social network industry, users, and the whole information industry tostudy the safety theory and key technologies.Usually, users register an account with their real information. Thus, the privacy data would be morelikely to be leaked such as loss of accounting information, ID theft and malicious software. As a result, howto protect the safety of users’information in social network will be and should be an important issue.This thesis mainly pays its attention to the structure of dynamic trust model in social networks.Considering the traditional and existing efforts, a study has been given on how to protect data privacy. Thestudy covers areas of network protocols, information safety, applying cryptology. The thesis will be showedin3parts as follows:(1) Deployment of dynamic trust model. On the basis of characters of social networks safety, a newdynamic trust model has been proposed considering the reputation and interaction between networks.RBtrust takes advantage of the most reputation information in the existing model to exact the relativeinformation from the interacting users thus making a side-weighted network to value the trust.(2) A dynamic trust model based on phone contacts. It is convenient to get access to the Internet viasmart phones. However, as for the openness of the wireless network, the safety is badly threatened. Thethesis fully employs the contact list in the phone and proposes an algorithm based on the contacts to get thereputation information, reducing the risk of leaking private data. This algorithm initially evaluates howhonest the members are in the contacts. According to the trusted members in the contacts, the algorithmcalculates the trust of unknown users. When the interaction is finished, it will update the trust degree.(3) The use of dynamic trust model in protecting the private data. After proposing the algorithm, ascheme has been put forward to help protect the privacy of users.Several experiments have been finished with PeerSim. The results show that the model couldpractically protect the users’ privacy in the networks and improve the success rate of interactions, makingcontributions to the healthy development of the social network.