动态社交网络中基于结构多样性的隐私保护方案

发布时间：2018-01-07 01:29

本文关键词：动态社交网络中基于结构多样性的隐私保护方案　出处：《湖南科技大学学报(自然科学版)》2016年04期 　论文类型：期刊论文

【摘要】：当前许多隐私保护技术主要针对静态社交网络.然而,由于数据连续发布,动态社交网络也存在隐私泄露问题.为了防范敌人的攻击,引入一种新的动态隐私保护方法,称为动态kw重结构多样性匿名法kw-SDA.该方法通过对个体分组保护,将连续发布数据时结点/社区身份的泄露概率限制为1/k.然后,提出一种可以实现动态kw-SDA算法的可拓展启发式算法.该算法可根据前w-1次发布的数据对图形进行匿名化处理,使图形改动最小化.此外,通过引入CS表,该算法可以逐渐汇总连续数据发布时的结点信息,避免了匿名化处理时扫描发布的所有数据.评估结果表明,该方法既能保护网络的大部分特征,又能有效保护隐私.
[Abstract]:At present, many privacy protection technologies are mainly aimed at static social networks. However, because of the continuous release of data, dynamic social networks also have privacy disclosure problems in order to guard against enemy attacks. A new dynamic privacy protection method, called dynamic kw multiplicity anonymity method, is introduced in this paper. Limit the probability of node / community identity disclosure to 1 / 1 / k. when data is published continuously. Then. This paper presents an extensible heuristic algorithm which can implement dynamic kw-SDA algorithm. The algorithm can process the graph anonymously according to the data released by the previous w-1, and minimize the changes of the graph. By introducing CS table, the algorithm can gradually aggregate the node information when continuous data is published, and avoid scanning all published data when anonymous processing. The evaluation results show that. This method can not only protect most of the features of the network, but also protect privacy effectively.
【作者单位】：辽宁机电职业技术学院信息工程系;
【基金】：院级科研基金项目(2014011)
【分类号】：TP309
【正文快照】： 先前的社交网络数据发布隐私保护技术主要针对静态社交网络.然而,随着社交网络数据的迅速发展,人们很少在某一时刻只发布一张快照并在发布之后对变化置若罔闻.随着动态社交网络分析和应用的需求不断上升,人们迫切需要解决动态社交网络数据连续发布时的隐私问题[1].图1给出了一

【参考文献】