大数据隐私保护中基于可信邻居选择的用户网络行为匿名技术研究  

Anonymity Technology of User Network Behavior Based on Trusted Neighbors Selecting for Big Data Privacy Protection

作　　者：周胜利[1,2] 陈光宣[2,3] 吴礼发[1] ZHOU Sheng-li CHEN Guang-xuan WU Li-fa(College of Command Information System, PLA University of Science and Technology, Nanjing 210007, China computer and Information Department, Zhejiang Police College, Hangzhou 310000, China Institute of Software Chinese Academy of Sciences, Beijing 100190, China)

机构地区：[1]解放军理工大学指挥信息系统学院,南京210007 [2]浙江警察学院计算机与信息技术系,杭州310000 [3]中国科学院软件研究所,北京100190

出　　处：《计算机科学》2016年第B12期136-139,151,共5页Computer Science

基　　金：本文受NSFC-浙江两化融合联合基金项目（U1509219）,国家高技术研究发展计划（“863”计划）基金资助项目（2015AA016003）资助.

摘　　要：传统的大数据用户网络行为隐私保护研究主要通过数据加密实现匿名访问，难以同时满足数据隐私保护所需要的不可追踪性及可信第三方审计所需要的可追踪性。针对该问题，设计了大数据中基于可信邻居选择的匿名方法。在可信任的第三方（主要指大数据政府管理机构）前提下，引入信任度随机邻居匿名机制，在邻近信任度区间随机选择邻居，利用盲签名技术对用户及随机邻居进行加密，模糊隐藏用户，防止恶意用户跟踪获取隐私，同时保证可信第三方机构对数据进行跟踪审计，挖掘潜在恶意网络行为。分析及实验表明，该方法在大数据平台上相对随机地采用随机假名匿名方法较节省存储空间，但效率稍低。The traditional research of privacy protection of user network behavior in big data mainly incorporates data encryption to achieve anonymous access, which is difficult to meet the needs of untraceability required by data privacy protection and traceability required by trusted third-party audits. To solve the problem, an anonymous method based on selecting trusted neighbors in big data was designed. On the premise of a trusted third party （mainly refers to big data government regulatory agencies）, a random trusted neighbor anonymous mechanism was introduced. Neighbors were selected randomly in the credit value interval, and blind signature technology was employed to encrypt users and random neighbors, vaguely hide users, prevent malicious users from tracing and acquiring privacy, while at the same time ensuring a trusted third party to trace and audit the data and tapping the potential malicious network behavior. Analysis and experiments have shown that this method relatively randomly selects random pseudonym and anonymous methods on a big data platform to save storage but has slightly low efficiency.

关 键 词：大数据 隐私保护 盲签名 信任度 身份签名 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]