Optimizing Batch Linear Queries under Exact and Approximate Differential Privacy
Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result such that it is provably hard for the adversary to infer the presence or absence of any individual record from the published...
Saved in:
| Published in | ACM transactions on database systems Vol. 40; no. 2; pp. 1 - 47 |
|---|---|
| Main Authors | , , , , , |
| Format | Journal Article |
| Language | English |
| Published |
01.06.2015
|
| Subjects | |
| Online Access | Get full text |
| ISSN | 0362-5915 1557-4644 |
| DOI | 10.1145/2699501 |
Cover
| Abstract | Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result such that it is provably hard for the adversary to infer the presence or absence of any individual record from the published noisy results. The main objective in differentially private query processing is to maximize the accuracy of the query results while satisfying the privacy guarantees. Previous work, notably Li et al. [2010], has suggested that, with an appropriate strategy, processing a batch of correlated queries as a whole achieves considerably higher accuracy than answering them individually. However, to our knowledge there is currently no practical solution to find such a strategy for an arbitrary query batch; existing methods either return strategies of poor quality (often worse than naive methods) or require prohibitively expensive computations for even moderately large domains. Motivated by this, we propose a low-rank mechanism (LRM), the first practical differentially private technique for answering batch linear queries with high accuracy. LRM works for both exact (i.e., ϵ-) and approximate (i.e., (ϵ, δ)-) differential privacy definitions. We derive the utility guarantees of LRM and provide guidance on how to set the privacy parameters, given the user's utility expectation. Extensive experiments using real data demonstrate that our proposed method consistently outperforms state-of-the-art query processing solutions under differential privacy, by large margins. |
|---|---|
| AbstractList | Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result such that it is provably hard for the adversary to infer the presence or absence of any individual record from the published noisy results. The main objective in differentially private query processing is to maximize the accuracy of the query results while satisfying the privacy guarantees. Previous work, notably Li et al. [2010], has suggested that, with an appropriate strategy, processing a batch of correlated queries as a whole achieves considerably higher accuracy than answering them individually. However, to our knowledge there is currently no practical solution to find such a strategy for an arbitrary query batch; existing methods either return strategies of poor quality (often worse than naive methods) or require prohibitively expensive computations for even moderately large domains. Motivated by this, we propose a low-rank mechanism (LRM), the first practical differentially private technique for answering batch linear queries with high accuracy. LRM works for both exact (i.e., ϵ-) and approximate (i.e., (ϵ, δ)-) differential privacy definitions. We derive the utility guarantees of LRM and provide guidance on how to set the privacy parameters, given the user's utility expectation. Extensive experiments using real data demonstrate that our proposed method consistently outperforms state-of-the-art query processing solutions under differential privacy, by large margins. Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result such that it is provably hard for the adversary to infer the presence or absence of any individual record from the published noisy results. The main objective in differentially private query processing is to maximize the accuracy of the query results while satisfying the privacy guarantees. Previous work, notably Li et al. [2010], has suggested that, with an appropriate strategy, processing a batch of correlated queries as a whole achieves considerably higher accuracy than answering them individually. However, to our knowledge there is currently no practical solution to find such a strategy for an arbitrary query batch; existing methods either return strategies of poor quality (often worse than naive methods) or require prohibitively expensive computations for even moderately large domains. Motivated by this, we propose a low-rank mechanism (LRM), the first practical differentially private technique for answering batch linear queries with high accuracy. LRM works for both exact (i.e., epsilon -) and approximate (i.e., ( epsilon , delta )-) differential privacy definitions. We derive the utility guarantees of LRM and provide guidance on how to set the privacy parameters, given the user's utility expectation. Extensive experiments using real data demonstrate that our proposed method consistently outperforms state-of-the-art query processing solutions under differential privacy, by large margins. |
| Author | Yang, Yin Winslett, Marianne Xiao, Xiaokui Hao, Zhifeng Yuan, Ganzhao Zhang, Zhenjie |
| Author_xml | – sequence: 1 givenname: Ganzhao surname: Yuan fullname: Yuan, Ganzhao organization: South China University of Technology, Guangzhou, China – sequence: 2 givenname: Zhenjie surname: Zhang fullname: Zhang, Zhenjie organization: Advanced Digital Sciences Center, Singapore – sequence: 3 givenname: Marianne surname: Winslett fullname: Winslett, Marianne organization: Advanced Digital Sciences Center and University of Illinois at Urbana-Champaign, IL – sequence: 4 givenname: Xiaokui surname: Xiao fullname: Xiao, Xiaokui organization: Nanyang Technological University, Singapore – sequence: 5 givenname: Yin surname: Yang fullname: Yang, Yin organization: Hamad Bin Khalifa University, Qatar – sequence: 6 givenname: Zhifeng surname: Hao fullname: Hao, Zhifeng organization: South China University of Technology and Guangdong University of Technology, China |
| BookMark | eNpl0L1OwzAUBWALFYm2IF7BGywBO7FjZyyl_EiVWiSYI8e5BqPUCbZTtTw9Qe0E01k-Hd17JmjkWgcIXVJyQynjt2leFJzQEzSmnIuE5YyN0JhkeZrwgvIzNAnhkxDCZCHGaL3qot3Yb-ve8Z2K-gMvrQPl8UsP3kLAvavB48VO6YiVq_Gs63y7sxsVAd9bY8CDi1Y1eO3tVun9OTo1qglwccwpentYvM6fkuXq8Xk-WyY6LWhMoKqU4kzJHAhJNUha1BIMoyBoBTzjsjJppoyoGSe8NowIIJkwktBKSiiyKbo-9A7nfPUQYrmxQUPTKAdtH0oqGBMkozIb6NWBat-G4MGUnR8e8PuSkvJ3s_K42SCTP1LbqKJtXfTKNv_8D3bfbsY |
| CitedBy_id | crossref_primary_10_1016_j_eswa_2024_125279 crossref_primary_10_1109_TKDE_2017_2697856 crossref_primary_10_1109_JIOT_2021_3052978 crossref_primary_10_1109_TBDATA_2017_2715334 crossref_primary_10_1016_j_future_2018_04_016 crossref_primary_10_1145_3589268 crossref_primary_10_1145_3626725 crossref_primary_10_14778_3594512_3594519 crossref_primary_10_1016_j_cose_2021_102529 crossref_primary_10_1007_s10619_018_07255_6 crossref_primary_10_1145_3299887_3299900 crossref_primary_10_14778_3467861_3467864 |
| Cites_doi | 10.1145/1835804.1835868 10.1137/S1052623497330963 10.5555/1953048.2021036 10.1137/040616413 10.1145/1989323.1989347 10.1145/773153.773173 10.14778/2350229.2350251 10.1137/050645506 10.1145/2213836.2213972 10.1016/j.dam.2007.02.013 10.1109/ICDE.2013.6544900 10.14778/1920841.1920970 10.1145/2448496.2448529 10.1145/2488608.2488652 10.1007/11761679_29 10.1088/0266-5611/28/11/115010 10.14778/2350229.2350252 10.1090/S0025-5718-97-00777-1 10.1145/1807085.1807104 10.1145/1807167.1807247 10.1007/s10208-009-9045-5 10.1007/11681878_14 10.1007/s00778-013-0309-y 10.1145/1374376.1374464 10.1145/1559795.1559812 10.1007/978-3-642-28914-9_18 10.1145/2213836.2213910 10.1016/S0167-6377(99)00074-7 10.1145/1835804.1835869 10.1137/09076828X 10.1109/FOCS.2010.85 10.1145/2463676.2465330 10.1137/080716542 10.1109/FOCS.2007.41 10.1109/ICDM.2009.11 10.14778/2350229.2350253 10.1109/FOCS.2010.12 10.1145/2046556.2046581 10.1145/1557019.1557090 10.1007/s12532-012-0044-1 10.1145/2213977.2214088 10.1093/imanum/drp035 10.1145/2068816.2068825 10.1145/1989323.1989348 10.1145/1390156.1390191 10.1145/102782.102783 10.14778/2168651.2168653 10.1109/ICDE.2012.48 10.1145/1806689.1806786 10.1145/1851182.1851199 10.29012/jpc.v4i1.612 10.1109/ICDE.2012.16 10.1145/2020408.2020487 10.1145/1806689.1806794 |
| ContentType | Journal Article |
| DBID | AAYXX CITATION 7SC 8FD JQ2 L7M L~C L~D |
| DOI | 10.1145/2699501 |
| DatabaseName | CrossRef Computer and Information Systems Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional |
| DatabaseTitle | CrossRef Computer and Information Systems Abstracts Technology Research Database Computer and Information Systems Abstracts – Academic Advanced Technologies Database with Aerospace ProQuest Computer Science Collection Computer and Information Systems Abstracts Professional |
| DatabaseTitleList | CrossRef Computer and Information Systems Abstracts |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Sciences (General) Computer Science |
| EISSN | 1557-4644 |
| EndPage | 47 |
| ExternalDocumentID | 10_1145_2699501 |
| GroupedDBID | --Z -DZ -~X .DC 23M 4.4 5GY 5VS 6J9 8US 8VB AAKMM AALFJ AAYFX AAYXX ABPPZ ACGFO ACGOD ACM ADBCU ADL ADMLS AEBYY AEFXT AEGXH AEJOY AEMOZ AENEX AENSD AETEA AFWIH AFWXC AHQJS AIAGR AIKLT AKRVB AKVCP ALMA_UNASSIGNED_HOLDINGS ASPBG AVWKF BDXCO CCLIF CITATION CS3 D0L EBS EJD FEDTE GUFHI HGAVV H~9 I07 IAO ICD IEA IGS IOF K1G LHSKQ N95 P1C P2P PQQKQ QWB RNS ROL RXW TAE TH9 U5U UPT WH7 X6Y XH6 XSW ZCA ZL0 7SC 8FD JQ2 L7M L~C L~D |
| ID | FETCH-LOGICAL-c291t-ebbaa54a86e002ce819d8ef41e71be5358bf23af7d4505df407e037f801b88e93 |
| ISSN | 0362-5915 |
| IngestDate | Thu Jul 10 19:20:54 EDT 2025 Thu Jul 03 08:15:38 EDT 2025 Thu Apr 24 22:56:19 EDT 2025 |
| IsDoiOpenAccess | false |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 2 |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c291t-ebbaa54a86e002ce819d8ef41e71be5358bf23af7d4505df407e037f801b88e93 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| OpenAccessLink | https://dr.ntu.edu.sg/bitstream/10356/81387/1/Optimizing%20Batch%20Linear%20Queries%20under%20Exact%20and%20Approximate%20Differential%20Privacy.pdf |
| PQID | 1744703183 |
| PQPubID | 23500 |
| PageCount | 47 |
| ParticipantIDs | proquest_miscellaneous_1744703183 crossref_primary_10_1145_2699501 crossref_citationtrail_10_1145_2699501 |
| PublicationCentury | 2000 |
| PublicationDate | 2015-06-01 |
| PublicationDateYYYYMMDD | 2015-06-01 |
| PublicationDate_xml | – month: 06 year: 2015 text: 2015-06-01 day: 01 |
| PublicationDecade | 2010 |
| PublicationTitle | ACM transactions on database systems |
| PublicationYear | 2015 |
| References | e_1_2_1_60_1 Hardt M. (e_1_2_1_28_1) e_1_2_1_20_1 e_1_2_1_41_1 e_1_2_1_45_1 e_1_2_1_62_1 e_1_2_1_22_1 e_1_2_1_43_1 e_1_2_1_64_1 e_1_2_1_49_1 e_1_2_1_26_1 e_1_2_1_47_1 Srebro N. (e_1_2_1_53_1); 17 Fiacco A. V. (e_1_2_1_24_1) 1968 Xiao X. (e_1_2_1_58_1) e_1_2_1_31_1 e_1_2_1_54_1 e_1_2_1_8_1 e_1_2_1_56_1 e_1_2_1_12_1 e_1_2_1_35_1 e_1_2_1_50_1 e_1_2_1_10_1 e_1_2_1_33_1 e_1_2_1_52_1 e_1_2_1_2_1 e_1_2_1_16_1 e_1_2_1_39_1 e_1_2_1_14_1 e_1_2_1_37_1 e_1_2_1_18_1 Ball K. (e_1_2_1_1_1) 1997; 31 e_1_2_1_42_1 Bertsekas D. P. (e_1_2_1_4_1) Billingsley P. (e_1_2_1_6_1) e_1_2_1_40_1 e_1_2_1_23_1 e_1_2_1_46_1 e_1_2_1_61_1 e_1_2_1_21_1 e_1_2_1_63_1 e_1_2_1_27_1 e_1_2_1_25_1 e_1_2_1_48_1 e_1_2_1_29_1 e_1_2_1_7_1 e_1_2_1_30_1 e_1_2_1_55_1 e_1_2_1_5_1 Nesterov Y. E. (e_1_2_1_44_1) e_1_2_1_57_1 e_1_2_1_3_1 e_1_2_1_13_1 e_1_2_1_34_1 e_1_2_1_51_1 e_1_2_1_11_1 e_1_2_1_32_1 e_1_2_1_17_1 e_1_2_1_38_1 e_1_2_1_15_1 e_1_2_1_36_1 e_1_2_1_59_1 e_1_2_1_9_1 e_1_2_1_19_1 |
| References_xml | – volume-title: Nonlinear Programming: Sequential Unconstrained Minimization Techniques year: 1968 ident: e_1_2_1_24_1 – volume-title: Nonlinear Programming ident: e_1_2_1_4_1 – ident: e_1_2_1_25_1 doi: 10.1145/1835804.1835868 – ident: e_1_2_1_7_1 doi: 10.1137/S1052623497330963 – ident: e_1_2_1_10_1 doi: 10.5555/1953048.2021036 – ident: e_1_2_1_11_1 doi: 10.1137/040616413 – ident: e_1_2_1_16_1 doi: 10.1145/1989323.1989347 – ident: e_1_2_1_17_1 doi: 10.1145/773153.773173 – ident: e_1_2_1_37_1 doi: 10.14778/2350229.2350251 – volume-title: Proceedings of the IEEE International Conference on Data Engineering (ICDE'10) ident: e_1_2_1_58_1 – ident: e_1_2_1_14_1 doi: 10.1137/050645506 – ident: e_1_2_1_46_1 doi: 10.1145/2213836.2213972 – ident: e_1_2_1_54_1 doi: 10.1016/j.dam.2007.02.013 – ident: e_1_2_1_47_1 doi: 10.1109/ICDE.2013.6544900 – ident: e_1_2_1_33_1 doi: 10.14778/1920841.1920970 – ident: e_1_2_1_36_1 doi: 10.1145/2448496.2448529 – ident: e_1_2_1_45_1 doi: 10.1145/2488608.2488652 – ident: e_1_2_1_19_1 doi: 10.1007/11761679_29 – ident: e_1_2_1_55_1 doi: 10.1088/0266-5611/28/11/115010 – ident: e_1_2_1_62_1 doi: 10.14778/2350229.2350252 – volume: 31 start-page: 1 year: 1997 ident: e_1_2_1_1_1 article-title: An elementary introduction to modern convex geometry publication-title: Flavors Geom. – ident: e_1_2_1_12_1 doi: 10.1090/S0025-5718-97-00777-1 – ident: e_1_2_1_34_1 doi: 10.1145/1807085.1807104 – ident: e_1_2_1_49_1 doi: 10.1145/1807167.1807247 – ident: e_1_2_1_9_1 doi: 10.1007/s10208-009-9045-5 – ident: e_1_2_1_21_1 doi: 10.1007/11681878_14 – ident: e_1_2_1_60_1 doi: 10.1007/s00778-013-0309-y – ident: e_1_2_1_8_1 doi: 10.1145/1374376.1374464 – ident: e_1_2_1_48_1 doi: 10.1145/1559795.1559812 – volume: 17 volume-title: Proceedings of the Conference on Advances Neural Information Processing Systems (NIPS'04) ident: e_1_2_1_53_1 – ident: e_1_2_1_15_1 doi: 10.1007/978-3-642-28914-9_18 – ident: e_1_2_1_61_1 doi: 10.1145/2213836.2213910 – volume-title: Introductory Lectures on Convex Optimization: A Basic Course ident: e_1_2_1_44_1 – ident: e_1_2_1_27_1 doi: 10.1016/S0167-6377(99)00074-7 – ident: e_1_2_1_5_1 doi: 10.1145/1835804.1835869 – ident: e_1_2_1_26_1 doi: 10.1137/09076828X – ident: e_1_2_1_30_1 doi: 10.1109/FOCS.2010.85 – ident: e_1_2_1_63_1 doi: 10.1145/2463676.2465330 – ident: e_1_2_1_2_1 doi: 10.1137/080716542 – ident: e_1_2_1_42_1 doi: 10.1109/FOCS.2007.41 – ident: e_1_2_1_32_1 doi: 10.1109/ICDM.2009.11 – ident: e_1_2_1_64_1 doi: 10.14778/2350229.2350253 – ident: e_1_2_1_22_1 doi: 10.1109/FOCS.2010.12 – volume-title: Probability and Measure ident: e_1_2_1_6_1 – ident: e_1_2_1_38_1 doi: 10.1145/2046556.2046581 – ident: e_1_2_1_41_1 doi: 10.1145/1557019.1557090 – ident: e_1_2_1_56_1 doi: 10.1007/s12532-012-0044-1 – ident: e_1_2_1_29_1 doi: 10.1145/2213977.2214088 – ident: e_1_2_1_3_1 doi: 10.1093/imanum/drp035 – ident: e_1_2_1_52_1 doi: 10.1145/2068816.2068825 – ident: e_1_2_1_57_1 doi: 10.1145/1989323.1989348 – ident: e_1_2_1_18_1 doi: 10.1145/1390156.1390191 – ident: e_1_2_1_23_1 doi: 10.1145/102782.102783 – ident: e_1_2_1_35_1 doi: 10.14778/2168651.2168653 – ident: e_1_2_1_39_1 – ident: e_1_2_1_59_1 doi: 10.1109/ICDE.2012.48 – ident: e_1_2_1_31_1 doi: 10.1145/1806689.1806786 – ident: e_1_2_1_40_1 doi: 10.1145/1851182.1851199 – ident: e_1_2_1_51_1 doi: 10.29012/jpc.v4i1.612 – ident: e_1_2_1_13_1 doi: 10.1109/ICDE.2012.16 – ident: e_1_2_1_43_1 doi: 10.1145/2020408.2020487 – volume-title: Proceedings of the Conference on Neural Information Processing Systems (NIPS'12) ident: e_1_2_1_28_1 – ident: e_1_2_1_20_1 doi: 10.1007/11761679_29 – ident: e_1_2_1_50_1 doi: 10.1145/1806689.1806794 |
| SSID | ssj0004897 |
| Score | 2.238895 |
| Snippet | Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into... |
| SourceID | proquest crossref |
| SourceType | Aggregation Database Enrichment Source Index Database |
| StartPage | 1 |
| SubjectTerms | Accuracy Approximation Privacy Queries Query processing Strategy Utilities |
| Title | Optimizing Batch Linear Queries under Exact and Approximate Differential Privacy |
| URI | https://www.proquest.com/docview/1744703183 |
| Volume | 40 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnV1Lb9QwELaW9sIFaAFRCshICIFQSp3YeRyXUqgQC0W00sIlspOJGmizaDdB1f56xo88tovE45K1ojjRZr7MjGc83xDyJIhFLgoIPMU4eJzlPo4i6YEqZJgUQcYNl97kQ3h0yt9NxXQ0Wg6rS2q1ly1_W1fyP1LFcyhXXSX7D5LtbooncIzyxSNKGI9_JeOP-L1flEu92n-FKlVTYFSamOdTo-mLF6bH7fzF4aWug9QB8rEmEL8s0UkF1HW2M0qtQ-bH8_KnzFYyvOODie4f0TYTN1kFvZ1Umz3H_9y5418aG0Z9i47mmZytBaO_nkH1reyzQGW1OHf7gye4VpdVn9ufltIEb_Xv96YcBiWY6DdPDYqxRGIrNffA6VYReTy0dI-t8rVcTQ5k_kCTsoFJtpyc68qea14MP0wS4Z69Qqd9xcx1mw9tKbZI3cRrZNOPIp3i3xy_nrz_3FfVxqY1T_dXbMm1nvrSTV31ZVZNufFPTm6RG25hQccWJVtkBNU2udk27aBOh2-TLTda0GeOcvz5bXLcI4kaJFGLJOqQRA2SqEESRSTRAZLoEEnUIekOOX1zeHJw5LleG17mJ6zGb1NJKbiMQ0AbmQE6inkMBWcQMQUiELEq_EAWUc7RZ84Lvh_BfhAV6OCoOIYkuEs2qlkF9whVOStEFsagJDrjwPCWyvdVjGsDQIPCdsjT9q2lmSOi1_1QztMrktkhtLvwh-VeWb_kcfvaU9SLOtklK5g1ixRX2jwyFuv-n2-zS673KH5ANup5Aw_R2azVI4eKXwLJgiA |
| linkProvider | EBSCOhost |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Optimizing+Batch+Linear+Queries+under+Exact+and+Approximate+Differential+Privacy&rft.jtitle=ACM+transactions+on+database+systems&rft.au=Yuan%2C+Ganzhao&rft.au=Zhang%2C+Zhenjie&rft.au=Winslett%2C+Marianne&rft.au=Xiao%2C+Xiaokui&rft.date=2015-06-01&rft.issn=0362-5915&rft.eissn=1557-4644&rft.volume=40&rft.issue=2&rft.spage=1&rft.epage=47&rft_id=info:doi/10.1145%2F2699501&rft.externalDBID=n%2Fa&rft.externalDocID=10_1145_2699501 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0362-5915&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0362-5915&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0362-5915&client=summon |