Research for Privacy Preserving in Social Networks based on GPGD

In the privacy issues of data mining for social networks, privacy issues of the edge of the weights between two nodes is particularly important. Although Greedy Perturbation algorithm (GP) for privacy protection can guarantee the shortest path between two specific nodes of social network unchanged, its length has a certain gap before and after the disturbance. Aimed at the deficiency, this paper proposes a Greedy Perturbation algorithm based on Gaussian Distribution (GPGD), which ensures that when the edge of the shortest path between two specific nodes of social network is disturbed, the shortest path between two nodes will not be changed, and the length of the shortest path will be closer than the length after the disturbance obtained by greedy perturbation algorithm. Verified by the analysis of simulation results, the proposed algorithm in this paper is effective.