Special Issue Article Open Access

Secure Personal Information by Analyzing Online Social Networks

Abstract

The information published in the social networks need to be elegant and more individualized. By recognizing this in social networks motivated us, to propose a scheme called privacy protection scheme which prevents the revelation of identities of both users and some selected features in their profiles. Each user can pick out the features of his own profile he wishes to hide. In this report, we simulate the users as nodes and the feature as labels in the social networks which are modeled as a graph .Labels in the graph are treated as sensitive or non-sensitive. The background knowledge held by the rivals and sensitive data or information that needed to be protected are considered or treated as node labels. We allow the graph data to be published in such a way that the rival who holds the information about node’s neighborhood cannot safely infer it’s both identity and its sensitive labels by presenting a privacy protection algorithm. This algorithm transforms the nodes in original graph as amply identical. The designed algorithm may lose little information but preserves its usefulness as much as it can. The original graph structure and its properties are also evaluated to find which extent the algorithms preserve privacy. We also demonstrated that the solution we proposed is effective, efficient and scalable than those in anterior research.

T. Satyendra kumar, C.Vijaya Rani

To read the full article Download Full Article | Visit Full Article