基于本地化差分隐私的数据立方体发布方法

张啸剑; 王浩锋; 曹小杰; 傅继彬

doi:10.11959/j.issn.2096-109x.2025048

您当前的位置：

首页 >

文章列表页 >

基于本地化差分隐私的数据立方体发布方法

学术论文 | 更新时间：2025-09-12

- 基于本地化差分隐私的数据立方体发布方法
- Data cubes publication under local differential privacy
- 网络与信息安全学报 2025年11卷第4期页码：141-159
- 作者机构：
  
  1.河南财经政法大学计算机与信息工程学院，河南郑州 450046
  2.河南省物联大数据处理与安全工程技术研究中心，河南郑州 450121
- 作者简介：
  
  [ "张啸剑（1980- ），男，河南周口人，博士，河南财经政法大学教授、硕士生导师，主要研究方向为差分隐私与机器学习、AI增强型数据管理、图数据管理。" ]
  [ "王浩锋（1999- ），男，河南许昌人，硕士，主要研究方向为差分隐私。" ]
  [ "曹小杰（1999- ），女，河南周口人，硕士，主要研究方向为差分隐私。" ]
  [ "傅继彬（1975- ），男，河南许昌人，博士，河南财经政法大学副教授、硕士生导师，主要研究方向为人工智能、计算机视觉。" ]
- 基金信息：
  
  国家自然科学基金(62072156);河南省自然科学杰出青年科学基金(252300421061);河南省重点科研项目计划基础研究专项(25ZX012)
- DOI：10.11959/j.issn.2096-109x.2025048
  中图分类号： TP393
- 收稿：2025-01-19，
  
  修回：2025-04-22，
  
  纸质出版：2025-08-25
- 稿件说明：
移动端阅览
张啸剑, 王浩锋, 曹小杰, 等. 基于本地化差分隐私的数据立方体发布方法[J]. 网络与信息安全学报, 2025, 11(4): 141-159.

Format: ZHANG X J, WANG H F, CAO X J, et al. Data cubes publication under local differential privacy[J]. Chinese Journal of Network and Information Security, 2025, 11(4): 141-159.
张啸剑, 王浩锋, 曹小杰, 等. 基于本地化差分隐私的数据立方体发布方法[J]. 网络与信息安全学报, 2025, 11(4): 141-159. DOI： 10.11959/j.issn.2096-109x.2025048.

Format: ZHANG X J, WANG H F, CAO X J, et al. Data cubes publication under local differential privacy[J]. Chinese Journal of Network and Information Security, 2025, 11(4): 141-159. DOI： 10.11959/j.issn.2096-109x.2025048.

摘要

基于本地化差分隐私的数据立方体发布已经得到研究者的广泛关注。针对现有方法在数据立方体发布过程中未能完全利用方体格结构的问题，提出了有效且满足本地差分隐私的数据立方体发布策略Full-Pub、Base-Pub和Partial-Pub。这3种策略利用格结构自身蕴含的偏序关系进行方体发布。为了有效利用方体格结构的特性，Full-Pub、Base-Pub和Partial-Pub分别从全方体、基本方体和部分方体的角度发布方体。其中Partial-Pub是在用户分组策略的基础上，利用贪心算法与集合覆盖技术在整个方体集合中寻找部分发布方体。Full-Pub、Base-Pub和Partial-Pub中的每个用户首先把自身数据进行本地编码，然后利用Cuboid-OUE、Cuboid-HHT和Cuboid-CMS对编码数据进行本地扰动，最后汇报给收集者。收集者结合每个用户的报告值重构噪音方体进行发布。在真实与合成数据集上进行实验，结果表明所提的数据立方体发布方法优于现有数据立方体发布方法。

Abstract

The release of data cubes based on local differential privacy has been widely studied by researchers. To address the problem that existing methods failed to fully utilize the lattice structure of cuboids during data cube publication

three efficient data cube publishing methods—Full-Pub

Base-Pub

and Partial-Pub—were proposed in this paper. All three algorithms made use of the partial order relationships among cuboids within the lattice structure for data cube publication. To effectively leverage the characteristics of the cuboid lattice

Full-Pub

Base-Pub

and Partial-Pub published the cuboids from the perspectives of the entire data cube

the base cuboid

and a part of the data cube

respectively. In particular

based on a user-grouping strategy

the Partial-Pub method integrated a greedy algorithm and set coverage to identify a subset of all cuboids. In Full-Pub

Base-Pub

and Partial-Pub

each user’s tuple was first encoded

and then perturbed using Cuboid-OUE

Cuboid-HHT

and Cuboid-CMS

respectively. Based on the reports collected from all users

the noise cuboids were reconstructed by the collector to answer OLAP queries. Extensive experiments were conducted on multiple real-world and synthetic datasets

and the results demonstrated the superiority of the proposed solutions.

关键词

Keywords

references

DUCHI J C , JORDAN M I , MARTIN J . Wainwright: local privacy and statistical minimax rates [Z ] . 2013: 429-438 .

CHATTERJEE S , DUCHI J C , LAFFERTY J D , et al . Local minimax complexity of stochastic convex optimization [C ] // NIPS , 2016 : 3423 - 3431 .

CORMODE G , KULKARNI T , SRIVASTAVA D . Marginal release under local differential privacy [C ] // SIGMOD , 2022 . 131 - 146 .

ZHANG Z K , WANG T H , LI N H , et al . CALM: Consistent adaptive local marginal for marginal release under local differential privacy [C ] // CCS , 2018 : 212 - 229 .

WARNER S L . Randomized response: a survey technique for eliminating evasive answer bias [J ] . Journal of the American Statistical Association , 1965 , 60 ( 309 ): 63 - 69 .

ERLINGSSON U , PIHUR V , KOROLOVA A . RAPPOR: randomized aggregatable privacy-preserving ordinal response [C ] // CCS , 2014 : 1054 - 1067 .

KAIROUZ P , BONAWITZ , K A , RAMAGE D . Discrete distribution estimation under local privacy [C ] // ICML , 2016 : 2436 - 2444 .

WANG T , BLOCI J , LI N , et al . Locally differentially private protocols for frequency estimation [C ] // Proc of the 26th USENIX Security Symposium (S&P) , 2017 : 729 - 745 .

DING B , WINSLETT M , HAN J W , et al . Differentially private data cubes: optimizing noise sources and consistency [C ] // SIGMOD Conference , 2011 : 217 - 228 .

DWORK C . Differential privacy [C ] // Proceedings of the 33rd International Colloquium on Automata, Languages and Programming (ICALP 2006) , Springer Verlag , 2006 .

DWORK , C , JING L . Differential privacy and robust statistics [C ] // Proceedings of the 41st Annual ACM Symposium on Theory of Computing (STOC 2009 ), 2009 : 71 - 380 .

DWORK C . Differential privacy in new settingss [C ] // Proceedings of the Twenty-First Annual ACM-SIAM Symposium on Discrete Algorithms (SODA 2010), Society for Industrial and Applied Mathematics , 2010 : 174 - 183 .

QARDAJI W , YANG W N , LI N H . PriView: practical differentially private release of marginal contingency tables [C ] // Proceedings of the 2014 ACM SIGMOD International Conference on Management of Data . New York : ACM , 2014 : 1435 - 1446 .

ZHANG Z K , WANG T H , LI N H , et al . PrivSyn: Differentially private data synthesis [C ] // USENIX Security Symposium , 2021 : 929 - 946 .

YANG M M , GUO T L , ZHU T Q , et al . Local differential privacy and its applications: A comprehensive survey [J ] . Computer Standards & Interfaces , 2024 , 89 : 103827 .

ACHARYA J , SUN Z T , ZHANG H Y . Hadamard response: estimating distributions privately, efficiently, and with little communication [C ] // AISTATS ， 2019 : 1120 - 1129 .

ACHARYA J , SUN Z T , ZHANG H Y . Communication efficient, sample optimal, linear time locally private discrete distribution estimation [J ] . arXiv preprint arXiv: 1802.04705 .

BARAK B , CHAUDHURI K , DWORK C , et al . Privacy, accuracy, and consistency too: a holistic solution to contingency table release [C ] // PODS , 2007 : 273 - 282 .

BALLE B , BELL J , GASC´ON A . The privacy blanket of the shuffle model C ] // CRYPTO , 2019 : 638 - 667 .

Differential Privacy Team . Learning with privacy at scale, Apple Inc [EB ] . 2017 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据