基于本地化差分隐私的时序位置发布方案研究  

Research on Time-Serial Location Data Publication Based on Local Differential Privacy

作　　者：康海燕[1] 冀源蕊 KANG Hai-yan;JI Yuan-rui(School of Information Management,Beijing Information Science and Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学信息管理学院信息安全系,北京100192

出　　处：《电子学报》2022年第9期2222-2232,共11页Acta Electronica Sinica

基　　金：国家社科基金(No.21BTQ079);教育部人文社科项目(No.20YJAZH046);国家自然科学基金(No.61370139)。

摘　　要：为了解决基于位置的服务(Location Based Service,LBS)在收集用户位置数据时造成的隐私泄露,提出一种本地化差分隐私位置发布模型.首先,该模型采用了灵活的位置隐私保护方案(个性化隐私设置),即由用户选择已设定的多种隐私策略或定制隐私策略,在此基础上设计了定制隐私策略位置扰动算法(Customized Privacy policy Lo⁃cation Perturbation algorithm,CPLP);其次,提出并设计一种基于隐马尔可夫模型的时序关联位置隐私发布算法(Tem⁃poral Relational Location Privacy publishing algorithm,TRLP),解决发布时序位置时产生的隐私泄露;最后,在GeoLife数据集和Gowalla数据集上通过对比实验验证了该模型的有效性.In order to solve the privacy leakage problem in location based service(LBS)when collecting user's loca⁃tion data,we proposed a time-serial location data publication model based on local differential privacy.Firstly,the model adapts a flexible location privacy preservation method,allows users to choose or customize their privacy policy(personal⁃ized privacy Settings),based on customized privacy policy,we designed a customized privacy policy location perturbation algorithm(CPLP);Secondly,we proposed and designed temporal relational location privacy publishing algorithm(TRLP)based on hidden Markov model(HMM),which can reduce the privacy leakage when releasing the time-serial location data.Finally,we verified the usability of the algorithm on data sets Geolife and Gowalla.

关 键 词：差分隐私 位置服务 时序数据 隐私发布 隐马尔可夫模型 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]