Towards Benchmarking Privacy Risk for Differential Privacy: A Survey

被引:0
|
作者
Prokhorenkov, Dmitry [1 ]
Cao, Yang [2 ]
机构
[1] Tech Univ Munich, Garching, Germany
[2] Hokkaido Univ, Sapporo, Hokkaido, Japan
关键词
data privacy; GDPR; differential privacy; privacy risk; attack; THREATS;
D O I
10.1145/3600100.3625373
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
While utilizing the Differential Privacy (DP) method, there are still many open issues regarding the method's effectiveness and reliability, including multiple difficulties in measuring the privacy level, which leads to implications in assessing risks while operating with personal data. Today, many experts and researchers emphasize the need and importance of utilizing DP for various kinds of data and tasks. However, to the best of our knowledge, the studies have yet to analyze the DP method regarding privacy risk assessment, thereby assessing the risks while utilizing DP and in the context of known types of privacy attacks. As a result, this study examines the existing privacy risks for various DP types in the context of present types of attacks. Also, the concept of privacy risk for DP will be analyzed, along with the corresponding metrics and their utility measurement. Our research method relies on a literature review; as a result, studies published from 2010 to 2023 were reviewed. Selected articles we examined based on the existing types of attacks, methods and metrics used to assess privacy risks. Based on this, we advanced the concept of privacy risk since none of the scientific studies clearly established the notation privacy risk for DP. Thus, we seek to explain the concept of privacy risk for DP in the context of existing types of attacks, thereby enabling DP utilization concerning the GDPR.
引用
收藏
页码:322 / 327
页数:6
相关论文
共 50 条
  • [41] Differential privacy in deep learning: Privacy and beyond
    Wang, Yanling
    Wang, Qian
    Zhao, Lingchen
    Wang, Cong
    FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, 2023, 148 : 408 - 424
  • [42] When Differential Privacy Implies Syntactic Privacy
    Ekenstedt, Emelie
    Ong, Lawrence
    Liu, Yucheng
    Johnson, Sarah
    Yeoh, Phee Lep
    Kliewer, Joerg
    IEEE Transactions on Information Forensics and Security, 2022, 17 : 2110 - 2124
  • [43] When Differential Privacy Implies Syntactic Privacy
    Ekenstedt, Emelie
    Ong, Lawrence
    Liu, Yucheng
    Johnson, Sarah
    Yeoh, Phee Lep
    Kliewer, Joerg
    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2022, 17 : 2110 - 2124
  • [44] Benchmarking Privacy Preserving Scientific Operations
    Aly, Abdelrahaman
    Smart, Nigel P.
    APPLIED CRYPTOGRAPHY AND NETWORK SECURITY, ACNS 2019, 2019, 11464 : 509 - 529
  • [45] Practical privacy-preserving benchmarking
    Kerschbaum, Florian
    PROCEEDINGS OF THE IFIP TC 11/ 23RD INTERNATIONAL INFORMATION SECURITY CONFERENCE, 2008, : 17 - 31
  • [46] Towards automated privacy risk assessments in IoT systems
    Markovic, Milan
    Asif, Waciar
    Corsar, David
    Jacobs, Naomi
    Edwards, Peter
    Rajarajan, Muttukrishnan
    Cottrill, Caitlin
    PROCEEDINGS OF THE 2018 WORKSHOP ON MIDDLEWARE AND APPLICATIONS FOR THE INTERNET OF THINGS (M4IOT '18), 2018, : 15 - 18
  • [47] Privacy Principles: Towards a Common Privacy Audit Methodology
    Makri, Eleni-Laskarina
    Lambrinoudakis, Costas
    TRUST, PRIVACY AND SECURITY IN DIGITAL BUSINESS, 2015, 9264 : 219 - 234
  • [48] Towards Spatial Range Queries Under Local Differential Privacy
    Zhang X.
    Fu N.
    Meng X.
    Jisuanji Yanjiu yu Fazhan/Computer Research and Development, 2020, 57 (04): : 847 - 858
  • [49] Towards A Differential Privacy and Utility Preserving Machine Learning Classifier
    Mivule, Kato
    Turner, Claude
    Ji, Soo-Yeon
    COMPLEX ADAPTIVE SYSTEMS 2012, 2012, 12 : 176 - 181
  • [50] Towards Private Learning on Decentralized Graphs With Local Differential Privacy
    Lin, Wanyu
    Li, Baochun
    Wang, Cong
    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2022, 17 : 2936 - 2946