Attacks on privacy and deFinetti's theorem

Kifer, Daniel

doi:10.1145/1559845.1559861

Cited by 136 publications

(131 citation statements)

References 39 publications

Supporting

Mentioning

128

Contrasting

Unclassified

Order By: Relevance

“…Some spectacular privacy breaches (such as demonstrations involving AOL [3] and GIC [39] data) have occurred when such intuition was not followed by a thorough analysis. In other cases, subtle implicit assumptions created weaknesses that could be exploited to breach privacy [25,40,22,26]. Similarly, the choice of a privacy mechanism based on some intuitively plausible measures of utility can result in a dataset that is not as useful as it could be [31,23].…”

Section: Introductionmentioning

confidence: 99%

A rigorous and customizable framework for privacy

Kifer

Machanavajjhala

2012

Proceedings of the 31st ACM SIGMOD-SIGACT-SIGAI Symposium on Principles of Database Systems

Self Cite

136

115

View full text Add to dashboard Cite

ABSTRACT"Privacy" and "utility" are words that frequently appear in the literature on statistical privacy. But what do these words really mean? In recent years, many problems with intuitive notions of privacy and utility have been uncovered. Thus more formal notions of privacy and utility, which are amenable to mathematical analysis, are needed. In this paper we present our initial work on an axiomatization of privacy and utility. In particular, we study how these concepts are affected by randomized algorithms. Our analysis yields new insights into the construction of both privacy definitions and mechanisms that generate data according to such definitions. In particular, it characterizes a class of relaxations of differential privacy and shows that desirable outputs of a differentially private mechanism are best interpreted as certain graphs rather than query answers or synthetic data.

show abstract

Section: Introductionmentioning

confidence: 99%

A rigorous and customizable framework for privacy

Kifer

Machanavajjhala

2012

Proceedings of the 31st ACM SIGMOD-SIGACT-SIGAI Symposium on Principles of Database Systems

Self Cite

136

115

View full text Add to dashboard Cite

show abstract

“…However, many of them have been shown to be insufficient due to realistic attacks on such schemes (e.g., see [3]). The notion of differential privacy [4,5], however, has remained strong and resilient to these attacks.…”

Section: Introductionmentioning

confidence: 99%

Crowd-Blending Privacy

Gehrke

Hay

Lui

et al. 2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract.We introduce a new definition of privacy called crowdblending privacy that strictly relaxes the notion of differential privacy. Roughly speaking, k-crowd blending private sanitization of a database requires that each individual i in the database "blends" with k other individuals j in the database, in the sense that the output of the sanitizer is "indistinguishable" if i's data is replaced by j's.We demonstrate crowd-blending private mechanisms for histograms and for releasing synthetic data points, achieving strictly better utility than what is possible using differentially private mechanisms. Additionally, we demonstrate that if a crowd-blending private mechanism is combined with a "pre-sampling" step, where the individuals in the database are randomly drawn from some underlying population (as is often the case during data collection), then the combined mechanism satisfies not only differential privacy, but also the stronger notion of zero-knowledge privacy. This holds even if the pre-sampling is slightly biased and an adversary knows whether certain individuals were sampled or not. Taken together, our results yield a practical approach for collecting and privately releasing data while ensuring higher utility than previous approaches.

show abstract

“…By fixing x and enumerating yi in DOM , the left hand side of Equation 16 can be regarded as a distribution fx(yi) of Yi. Given the theoretical values on the right hand side of Equation 16, the error of fx(yi) can again be quantified by the L2 norm: Failure of independent perturbation.…”

Section: Methodsmentioning

confidence: 99%

“…There have been a plethora of studies on partition-based approaches, such as generalization [28], anatomy [33,35], condensation [1]. Despite the popularity of these approaches, they are vulnerable to various types of privacy attacks, as pointed out in [16,26,32].…”

Section: Related Workmentioning

confidence: 99%

Optimal random perturbation at multiple privacy levels

2009

View full text Add to dashboard Cite

Random perturbation is a popular method of computing anonymized data for privacy preserving data mining. It is simple to apply, ensures strong privacy protection, and permits effective mining of a large variety of data patterns. However, all the existing studies with good privacy guarantees focus on perturbation at a single privacy level. Namely, a fixed degree of privacy protection is imposed on all anonymized data released by the data holder. This drawback seriously limits the applicability of random perturbation in scenarios where the holder has numerous recipients to which different privacy levels apply.Motivated by this, we study the problem of multi-level perturbation, whose objective is to release multiple versions of a dataset anonymized at different privacy levels. The challenge is that various recipients may collude by sharing their data to infer privacy beyond their permitted levels. Our solution overcomes this obstacle, and achieves two crucial properties. First, collusion is useless, meaning that the colluding recipients cannot learn anything more than what the most trustable recipient (among the colluding recipients) already knows alone. Second, the data each recipient receives can be regarded (and hence, analyzed in the same way) as the output of conventional uniform perturbation. Besides its solid theoretical foundation, the proposed technique is both space economical and computationally efficient. It requires O(n+m) expected space, and produces a new anonymized version in O(n +log m) expected time, where n is the cardinality of the original dataset, and m the number of versions released previously. Both bounds are optimal under the realistic assumption that n m.

show abstract

Attacks on privacy and deFinetti's theorem

Cited by 136 publications

References 39 publications

A rigorous and customizable framework for privacy

A rigorous and customizable framework for privacy

Crowd-Blending Privacy

Optimal random perturbation at multiple privacy levels

Contact Info

Product

Resources

About