2014 | OriginalPaper | Buchkapitel
Differentially Private Data Release: Improving Utility with Wavelets and Bayesian Networks
verfasst von : Xiaokui Xiao
Erschienen in: Web Technologies and Applications
Verlag: Springer International Publishing
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Privacy-preserving data publishing is an important problem that has been the focus of extensive study. The state-of-the-art privacy model for this problem is
differential privacy
, which offers a strong degree of privacy protection without making restrictive assumptions about the adversary. In this paper, we review two methods,
Privelet
and
PrivBayes
, for improving utility in differentially private data publishing.
Privelet
utilizes wavelet transforms to ensure that any range-count query can be answered with noise variance that is polylogarithmic to the size of the input data domain. Meanwhile,
PrivBayes
employs Bayesian networks to publish high-dimensional datasets without incurring prohibitive computation overheads or excessive noise injection.