基于属性分割的差分隐私高维数据发布方法

打开文本图片集
DOI:10.16652/j.issn.1004-373x.2025.18.018 引用格式:,.基于属性分割的差分隐私高维数据发布方法[J].现代电子技术,2025,48(18):114-118.
关键词:属性分割;差分隐私算法;高维数据;数据发布;贝叶斯网络;隐私预算;敏感度中图分类号:TN911.23-34;TP391.4 文献标识码:A 文章编号:1004-373X(2025)18-0114-05
Method of DP high dimensional data dissemination based on attribute segmentation
ZHANGHeng,HU Wanru (LiaoningUniversity,Shenyang11Oo36,China)
Abstract:Inordertoaddressthephenomenonof decreaseddataavailabilityduring high-dimensional datadisemination,a methodofdiferentialprivacyhigh-dimensionaldatadisseminationisproposed.Thismethdisbasedonatributesegntation technologyandguidedbydiferentialprivacy(DP)algorithm.Thehigh-dimensionaldatasetsaredividedbasedonattribute sensitivity,andteinformationeroyandmaxiuminformationcoefiientaresedtoidentifthesensitivityfarbtesand thedependencebetweenatributes.Aordingtothesensitivityofthedataset,itisdividedintothreetypes:sensitive,general sensitive,andlowsensitivedatasets.The Bayesiannetwork with DPiscreated,andthenoise-adingconditionsforprivacy budgetaregeneratedtocompletethesecuresharingofhigh-dimensionaldataunderdiferentialprivacymechanism.The experimentalresultsshowthattheproposedmethodcannotonlyefectivelycontroltheriskof privacyleakage,butalsomaintain theavailability of data well.
Keywords:atributesegmentation;diferential privacy;highdimensional data;datadissemination;Bayesiannetwork; privacy budget;sensitivity
0 引言
高维数据在发布与共享的过程中,会存在隐私泄露的问题,亟需在保护数据隐私的同时,又可以将这些高维数据充分利用起来,因此差分隐私(DP)算法被广泛研究和使用。(剩余6174字)