面向加权社会网络最短路径的隐私保护方法研究

发布时间：2018-01-06 03:04

本文关键词：面向加权社会网络最短路径的隐私保护方法研究　出处：《广西大学》2016年硕士论文　论文类型：学位论文

【摘要】：社会网络上的数据规模伴随着互联网技术发展的步伐不断增大,如此巨量的数据给研究人员进行知识挖掘以及数据分析工作提供便利的同时,隐藏在社会网络数据下的个人隐私信息的保护问题也随之而来。然而在加权社会网络发布之前,数据若未经处理则可能造成用户隐私泄露。需要注意的是,最短路径通常包含大量隐私信息,因此针对加权社会网络上的最短路径的保护显得尤为重要。基于此,本文的研究内容主要针对如何保护加权社会网络中最短路径上的权重信息和路径信息,本文主要工作如下：1)针对敏感节点对最短路径上的边权重泄露问题,提出了基于高斯分布的贪心扰动算法(Greedy Perturbation algorithm based on Gaussian Distribution, GPGD)。该算法不仅能够防止敏感节点对最短路径上边权重泄露,而且利用高斯分布的特性,有效改善了传统贪心扰动算法中敏感节点对最短路径长度在扰动前后存在较大差距问题；2)针对敏感节点对最短路径泄露问题,提出了top-k路径匿名隐私保护算法(Top-K-Single Path Anonymization Algorithm, TKSP)。该算法不仅能够防止敏感节点对最短路径泄露,还能保证扰动前后图数据特性一致,并且与KSP算法相比具有运行时间短、图数据保持性高及查询准确率高的优点。
[Abstract]:The scale of data on the social network is increasing with the development of Internet technology. Such a large amount of data provides convenience for researchers to do knowledge mining and data analysis at the same time. The protection of personal privacy information hidden under the social network data also follows. However, before the weighted social network is released, the unprocessed data may cause user privacy disclosure. The shortest path usually contains a lot of privacy information, so it is very important to protect the shortest path on the weighted social network. This paper mainly focuses on how to protect the weight information and path information of the shortest path in the weighted social network. The main work of this paper is as follows: 1) aiming at the leakage of the edge weight of the sensitive node to the shortest path. A greedy perturbation algorithm based on Gao Si distribution is proposed. Greedy Perturbation algorithm based on Gaussian Distribution. The algorithm can not only prevent sensitive nodes from leakage of edge weights on the shortest path, but also make use of the characteristics of Gao Si distribution. It effectively improves the gap between the length of the shortest path and the sensitive node in the traditional greedy perturbation algorithm before and after disturbance. 2) aiming at the shortest path leakage problem of sensitive nodes. An anonymous privacy protection algorithm for top-k path named Top-K-Single Path Anonymization Algorithm is proposed. The algorithm not only can prevent sensitive nodes from leaking the shortest path, but also can ensure the same characteristics of graph data before and after disturbance, and has shorter running time than KSP algorithm. The advantages of high graph data retention and high query accuracy.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP393.08

【相似文献】