Research On Privacy Protection Model Of Social Network Based On Sharing Habit

With the emergence of social networking tools such as We Chat and Twitter,online social networks have become an important channel for people to communicate and obtain information in their daily lives,but they have also become a major disaster area for unscrupulous companies and hackers to collect users’ private data.Existing research on the privacy protection of social networks mainly considers individuals’ privacy in social networks from the perspective of users as the research entity,and there is a lack of research on the privacy leakage problem caused by sharing information behaviors of users between communities such as "groups".As users join more and more groups with clear community labels,a piece of information posted by a member meant to be shared within a community may be leaked by other members to outside parties through crosscommunity interactions,which in effect increases the risk of privacy leakage.At the same time,if privacy settings are too strict,users’ usage of social networks can be greatly restricted,reducing the availability of social networks,which is contrary to the original intent of increasing user influence and strengthening friendships.Based on users’ sharing habits,attribute characteristics,and trust,this thesis designs a privacy protection model applicable to information sharing between communities.Considering the problem of promoting information sharing among users while protecting the privacy information shared by members within the same community,the thesis conducts research on both sharing-habits based privacy control methods between communities and trust-based privacy-protection information sharing between communities,as follows:(1)Aiming at the privacy leakage problem of shared information within communities that is normally ignored in online social networks,the thesis proposes an inter-community privacy control scheme based on sharing habits to reduce the privacy leakage risk of information shared in a community.To begin with,we formularize the credibility of users based on their attributes within a community.Then,considering the behavioral patterns of sharing interactive information between users,i.e.sharing habits,we measure the probability of users sharing information across communities according to the number of shared interactions between users,which can quantify their ability to keep the information confidential.Finally,combining user credibility and sharing probability,we design two blocking algorithms to control information flow between communities and reduce the risk of information flowing to adversaries outside the community.The simulation analysis indicates the effectiveness of the proposed framework in protecting information privacy in a community,and the privacy control effect of edge blocking is better than that of node blocking.(2)Aiming at the problem that users who are authorized to access information in a community want to share information with their friends across the community,but cannot guarantee the privacy security of users’ information in the community,we propose a trust-based privacy-protected information sharing mechanism between communities.Firstly,the potential privacy loss caused by the recipient is measured based on information sensitivity and user trust.Secondly,share rules are designed for whether to share information with friends in another community by combining privacy loss and user trust.Finally,this thesis trade-off between the privacy protection of information within a community and information sharing with friends based on user feedback.Simulation analysis shows that a trust-based information sharing mechanism can help reduce privacy loss while providing users with good information sharing in social networks.