面向社交网络数据的等差数列聚类匿名算法

Anonymity Algorithm of Arithmetic Sequence Clustering for Social Network Data

针对社交网络隐私保护如何减少信息损失,实现数据可用性的问题,提出一种个性化等差数列聚类匿名分配算法(PAS-CAA)。首先对选取的初始节点进行优化,基于综合相似度进行聚类,使每个超点至少包含k个节点;区分非敏感超点集和敏感超点集,对敏感超点集采用递减等差数列进行聚类,灵活地调节保护力度,对非敏感超点集实现基本的k保护力度;最后对超点进行匿名化处理。仿真实验结果表明算法在保护社交网络用户隐私的同时可以减少信息的损失,保留统计属性,实现了社交网络的个性化隐私保护。

Aiming at the problem of how to reduce information loss and achieve data availability in social network privacy protection,a personalized arithmetic sequence clustering anonymous allocation algorithm(PAS-CAA)was proposed.Firstly,it optimized the selection of initial nodes,cluster based on comprehensive similarity so that each super point contained at least k nodes.Then it distinguished between non-sensitive super point sets and sensitive super point sets,and used decreasing arithmetic sequence for sensitive super point sets to cluster,flexibly adjusted the protection intensity,realized the basic k protection intensity for the non-sensitive super-point set.Finally,it anonymized the super-points.The simulation experiment results showed that while protecting the privacy of social network users,the algorithm could reduce the loss of information,retain statistical attributes,and realize the personalized privacy protection of social networks.