Research On The Techniques For Privacy Preserving On Shortest Path Of Weighted Social Networks

The size of the data on the social network along with the development of Internet technology becomes bigger and bigger, such a huge amount of data provide convenience to the researchers for data analysis and data mining at the same time, data protection of personal privacy information hidden under the social network followed. However, publishing raw data may leak the user privacy. It is important to note that the shortest path usually contains a large number of privacy information, so the protection of weighted shortest path on social network is particularly significance. Based on this, how to protect the weights information and path information on the shortest path in the weighted social network becomes the research focus of this article, in this paper, the main work is as follows:1) For the edge weight of shortest path leak problem between sensitive nodes, we propose Greedy perturbation algorithm based on Gaussian distribution (GPGD). This algorithm uses the characteristics of guassian distribution, effectively improves problem that there is a big gap between sensitive nodes before and after the disturbance in the traditional greedy perturbation. And this algorithm can guarantee that the shortest path is changeless, and the length of the shortest path after the disturbance is closer to before.2) For the shortest path leak problem between sensitive nodes, we propose Top-K-Single Path Anonymization Algorithm (TKTP), this algorithm constructs k-1 paths which length are equal to the length of shortest path, and an attacker can’t identify the shortest path with the probability of more than 1/k. It can not only achieve the aim to protect the privacy of the shortest path, but also solve the problem of figure data inconsistency in existing algorithms, and has a short operation time, high figure data retention and query the advantage of high accuracy.