Answering Spatial Density Queries Under Local Differential Privacy

被引:1
作者
Tire, Ekin [1 ]
Gursoy, M. Emre [1 ]
机构
[1] Koc Univ, Dept Comp Engn, TR-34450 Istanbul, Turkiye
关键词
Protocols; Differential privacy; Sensitivity; Privacy; Servers; Estimation; Spatial databases; Data privacy; geospatial data analysis; local differential privacy (LDP); location privacy; spatial crowdsourcing; RANGE QUERIES; NP-COMPLETENESS;
D O I
10.1109/JIOT.2024.3357570
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Spatial density queries are fundamental in many geospatial data analysis and crowdsourcing tasks. However, answering spatial density queries may violate users' privacy by exposing their locations to an untrusted data collector. In this article, we propose a solution for answering spatial density queries under local differential privacy (LDP), a state-of-the-art privacy protection standard. Our solution consists of four main steps: 1) partitioning; 2) finding sensitivity; 3) user-side noisy response computation; and 4) server-side estimation. For the first step, we propose and analyze three basic partitioning strategies, and based on our analysis, we design an improved strategy called Advanced Partitioning. For the second step, we adapt graph-based modeling of query sets from the centralized differential privacy literature. Advanced Partitioning also leverages and extends this technique by formulating the partitioning problem using vertex coloring. For the third and fourth steps, in addition to adapting two popular LDP protocols (GRR and RAPPOR), we propose a novel extension for the optimized unary encoding protocol. Our new protocol (OBE) is not only applicable to our problem but can also be used in other LDP problems with bitvector encodings. Finally, we perform an extensive experimental evaluation of different partitioning strategies and protocols using multiple real-world data sets. Results show that Advanced Partitioning and OBE yield the lowest error, demonstrating the superiority of our proposed methods.
引用
收藏
页码:17419 / 17436
页数:18
相关论文
共 66 条
[1]   Building Quadtrees for Spatial Data Under Local Differential Privacy [J].
Alptekin, Ece ;
Gursoy, M. Emre .
DATA AND APPLICATIONS SECURITY AND PRIVACY XXXVII, DBSEC 2023, 2023, 13942 :22-39
[2]  
[Anonymous], Taxi trajectory prediction dataset: ECML/PKDD
[3]  
Apple Inc. Cupertino CA USA, 2020, Learning With Privacy at Scale
[4]   Local Differential Privacy for Deep Learning [J].
Arachchige, Pathum Chamikara Mahawaga ;
Bertok, Peter ;
Khalil, Ibrahim ;
Liu, Dongxi ;
Camtepe, Seyit ;
Atiquzzaman, Mohammed .
IEEE INTERNET OF THINGS JOURNAL, 2020, 7 (07) :5827-5842
[5]  
Arcolezi H. H., Digit. Commun. Netw.
[6]  
Arcolezi H.H., 2023, EDBT 2023, P512, DOI [DOI 10.48786/EDBT.2023.44, 10.48786/edbt.2023.44]
[7]  
Bassily R., 2017, ADV NEURAL INFORM PR, V30, P2288, DOI DOI 10.5555/3294771.3294989
[8]   Local, Private, Efficient Protocols for Succinct Histograms [J].
Bassily, Raef ;
Smith, Adam .
STOC'15: PROCEEDINGS OF THE 2015 ACM SYMPOSIUM ON THEORY OF COMPUTING, 2015, :127-135
[9]  
Chen R, 2016, PROC INT CONF DATA, P289, DOI 10.1109/ICDE.2016.7498248
[10]  
Cho E., 2011, PROC 17 ACM SIGKDD I, P1082