Preventing and Detecting Infiltration on Online Social Networks

Abstract

Nowadays, together with the development of internet, online social networks provide lots of benefits for human. People use social networks for different purposes, such as: communicating, information sharing, relations creating or for business purposes, and etc. However, accompanying with benefits of social networks, users also must face the security and privacy risks. These issues have recently paid much attention to. One of these issues is the user can penetrate and steal personal information. The assailants can penetrate through agreeing their friend requests. When confirming this friend request, the user unintentionally discloses personal information. Especially, if the user stolen information is an important person of a specific organization, losses are extremely considerable. Promoted by this phenomenon, in this paper, we propose a new solution to prevent and discover any penetration for a specific user in an organization. First of all, we propose a new model which is called as safety community model in order to protect everybody in the organization. We build a target function orienting to the safety for everybody in the organization. After that, we have designed an effective algorithm to discover the penetration of unsafe factors for specific users in the organization. Tests in social networks are actually implemented, and the result shows that our model can prevent the penetration of outside objects.