基于向量相似的权重社会网络隐私保护  

Privacy Preserving Based on Vector Similarity for Weighted Social Networks

作　　者：兰丽辉[1,2] 鞠时光[1] 

机构地区：[1]江苏大学计算机科学与通信工程学院,江苏镇江212013 [2]沈阳大学信息工程学院,辽宁沈阳110000

出　　处：《电子学报》2015年第8期1568-1574,共7页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61003288;61111130184);国家教育部博士点基金(No.20093227110005);江苏省普通高校研究生科研创新计划项目(No.CX10B_006X)

摘　　要：针对权重社会网络发布,提出采用基于向量相似的随机扰动方法实现多个发布场景下网络结构和边权重的隐私保护.该方法以边空间理论为基础,采用基于节点聚类的分割方法构建权重社会网络的向量集模型;以加权欧氏距离作为向量相似的度量标准,根据选定阈值构建发布候选集;从候选集随机选取向量实现权重社会网络的发布;可抵御多种节点识别攻击,迫使攻击者在一个向量发生概率相同的庞大结果集中进行重识别,增加了识别的不确定性.实验结果表明,该方法在确保社会个体隐私安全同时可保护社会网络分析所需的某些结构特征,提高发布数据效用.Aiming at the publication of weighted social networks,a random perturbation method based on vector similarity is proposed. It can protect network structures and edge weights in multiple release scenarios. It constructs vector set models by segmentation based on vertex cluster using edge space theory. It adopts weighted Euclidean distance as similarity metrics to construct the released candidate sets according to the threshold. It randomly selects vectors from candidate sets to construct the published weighted social networks. The proposed method can resist multiple vertex recognition attacks, force attackers to re-identify in a large result set that the existential probabilities of the vectors are same, and increase the uncertainty of recognition. The experimental results demonstrate that it can preserve individuals＇ privacy security, meanwhile it can protect some structure characteristics for networks analysis and improve data utility.

关 键 词：社会网络 边权重 隐私保护 向量集模型 加权欧氏距离 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]