Understanding the Sparse Vector Technique for Differential Privacy

Summary: Analyzes SVT variants for differential privacy, exposing flaws and misunderstandings across interactive and non-interactive settings. Proposes a tighter SVT with improved utility; in non-interactive DP, EM outperforms SVT, while interactive gains vary. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: 11747
Venue: VLDB
Year: 2017
Pagerank: 9.3073552e-05
Overall Rank: 2,012 | 86.20%
DOI: 10.14778/3055330.3055331

Incoming Non-self Citations Over Time

Authors

1. Min Lyu (University of Science and Technology Beijing)
2. Dong Su (Purdue University)
3. Ninghui Li (Purdue University)

BibTeX Citation

@article{lyu_vldb17,
        title = {{Understanding the Sparse Vector Technique for Differential Privacy}},
        author = {Lyu, Min and Su, Dong and Li, Ninghui},
        journal = {PVLDB},
        series = {{VLDB} '17},
        volume = {10},
        number = {6},
        pages = {637--648},
        doi = {10.14778/3055330.3055331},
        url = {https://doi.org/10.14778/3055330.3055331},
        year = {2017}
}

Incoming Citations (Sorted by Pagerank)

Showing 12 of 12 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
2,436	Computing Local Sensitivities of Counting Queries with Joins	2020	SIGMOD	8.5806427e-05
2,818	CryptEpsilon: Crypto-Assisted Differential Privacy on Untrusted Servers	2020	SIGMOD	8.0932359e-05
6,804	A Neural Database for Differentially Private Spatial Range Queries	2022	VLDB	5.768025e-05
7,338	DPGraph: A Benchmark Platform for Differentially Private Graph Analysis	2021	SIGMOD	5.6393827e-05
7,831	Differentially Private Hierarchical Heavy Hitters	2024	PODS	5.5357919e-05
8,908	Privacy and Accuracy-Aware AI/ML Model Deduplication	2025	SIGMOD	5.3483178e-05
9,054	Unleash the Power of Ellipsis: Accuracy-enhanced Sparse Vector Technique with Exponential Noise	2025	VLDB	5.3251649e-05
9,303	SPAS: Continuous Release of Data Streams under w-Event Differential Privacy	2025	SIGMOD	5.289545e-05
9,569	Free Gap Information from the Differentially Private Sparse Vector and Noisy Max Mechanisms	2020	VLDB	5.2528121e-05
9,602	Answering Private Linear Queries Adaptively using the Common Mechanism	2023	VLDB	5.2487195e-05
10,150	Concurrent Composition for Differentially Private Continual Mechanisms	2026	PODS	5.093636e-05
11,349	DP-PQD: Privately Detecting Per-Query Gaps In Synthetic Data Generated By Black-Box Mechanisms	2024	VLDB	5.093636e-05

Outgoing Citations (Sorted by Pagerank)

Showing 3 of 3 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
123	Revealing Information while Preserving Privacy	2003	PODS	0.00031082693
1,169	PrivBayes: Private Data Release via Bayesian Networks	2014	SIGMOD	0.00011838753
1,308	PrivTree: A Differentially Private Algorithm for Hierarchical Decompositions	2016	SIGMOD	0.00011216361

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	6,441	Differential Privacy in Data Publication and Analysis	2012	SIGMOD
2	9,602	Answering Private Linear Queries Adaptively using the Common Mechanism	2023	VLDB
3	9,086	Privacy Preservation by Disassociation	2012	VLDB
4	10,335	A General Framework for Per-record Differential Privacy	2026	SIGMOD
5	11,203	Personalized Truncation for Personalized Privacy	2024	SIGMOD
6	3,232	Privacy-preserving Anonymization of Set-valued Data	2008	VLDB
7	2,108	An Adaptive Mechanism for Accurate Query Answering under Differential Privacy	2012	VLDB
8	2,856	Publishing Set-Valued Data via Differential Privacy	2011	VLDB
9	9,569	Free Gap Information from the Differentially Private Sparse Vector and Noisy Max Mechanisms	2020	VLDB
10	9,054	Unleash the Power of Ellipsis: Accuracy-enhanced Sparse Vector Technique with Exponential Noise	2025	VLDB