|
gptkbp:instanceOf
|
Privacy Technique
|
|
gptkbp:application
|
Machine learning
Data analysis
Census data release
|
|
gptkbp:author
|
gptkb:Adam_Smith
gptkb:Cynthia_Dwork
gptkb:Frank_McSherry
gptkb:Kobbi_Nissim
|
|
gptkbp:category
|
Privacy-preserving data analysis
|
|
gptkbp:challenge
|
Balancing privacy and data utility
Composability of privacy guarantees
Parameter selection
|
|
gptkbp:field
|
Computer Science
Statistics
Data Privacy
|
|
gptkbp:guarantees
|
Indistinguishability of outputs
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Differential Privacy
|
|
gptkbp:impact
|
Widely adopted in academia and industry
|
|
gptkbp:influenced
|
gptkb:Google_Chrome
gptkb:Microsoft
gptkb:Uber
gptkb:Apple_iOS
|
|
gptkbp:introduced
|
gptkb:Cynthia_Dwork
|
|
gptkbp:introducedIn
|
2006
|
|
gptkbp:key
|
gptkb:Epsilon
Delta
|
|
gptkbp:limitation
|
Utility-privacy tradeoff
|
|
gptkbp:mathematicalDefinition
|
Probability of output changes little when one individual's data is changed
|
|
gptkbp:method
|
Adding random noise to data or queries
|
|
gptkbp:property
|
Mathematically quantifies privacy loss
|
|
gptkbp:publicationYear
|
2006
|
|
gptkbp:publishedIn
|
Calibrating Noise to Sensitivity in Private Data Analysis
|
|
gptkbp:purpose
|
Protect individual privacy in statistical databases
|
|
gptkbp:relatedConcept
|
gptkb:k-anonymity
gptkb:l-diversity
gptkb:t-closeness
|
|
gptkbp:relatedStandard
|
gptkb:NIST_SP_800-188
|
|
gptkbp:relatedTo
|
gptkb:Machine_Learning
gptkb:Data_Security
Cryptography
|
|
gptkbp:standardizedBy
|
gptkb:NIST
|
|
gptkbp:supportsAlgorithm
|
gptkb:Exponential_Mechanism
gptkb:Gaussian_Mechanism
gptkb:Laplace_Mechanism
|
|
gptkbp:type
|
gptkb:Global_Differential_Privacy
gptkb:Local_Differential_Privacy
|
|
gptkbp:usedBy
|
gptkb:US_Census_Bureau
|
|
gptkbp:bfsParent
|
gptkb:Cynthia_Dwork
|
|
gptkbp:bfsLayer
|
4
|