Abstract: With the widespread popularity of social network platforms, privacy disclosure has become a focus when users share personal information. There is a lack of unified evaluation on the privacy preserving methods for noisy social network. From the perspective of privacy mining, we choose the wiener filtering of signal processing which can automatically suppress and eliminate noise on the neighbor graph, and without considering background knowledge. Then we propose a social network privacy preserving intensity evaluation model, and design a privacy protection intensity evaluation algorithm. To verify the feasibility of the algorithm, this paper uses statistical properties of the undirected graph to measure privacy preserving intensity. The experiments conducted also show that the evaluation method can filter out some noise in the neighbor graph, achieve the purpose of the privacy preserving intensity evaluation, and provide guidance for theoretical research on privacy preserving of social network.

Key words: social network, privacy preserving, neighbor graph, privacy mining, wiener filtering

摘要： 随着社交网络平台的广泛普及,用户在分享个人信息时引起的隐私泄露问题成为人们关注的焦点.针对加噪型的社交网络隐私保护方法缺乏统一性评价的问题,本文在不考虑背景知识的前提下,从隐私挖掘的角度,选择信号处理中能够自动抑制噪声的维纳滤波对邻近图进行去噪,并且提出社交网络隐私保护强度评价模型,设计隐私保护强度评价算法.为了验证算法的可行性,论文采用无向图统计特征进行度量.实验表明,评估方法能够滤除邻近图中的部分噪声,达到了隐私保护强度评价的目的,也为社交网络隐私保护的理论研究提供了指导.

关键词: 社交网络, 隐私保护, 邻近图, 隐私挖掘, 维纳滤波