The Algorithmic Foundations of Differential Privacy

Dwork, C., & Roth, A. (2014). The Algorithmic Foundations of Differential Privacy. Foundations and Trends in Theoretical Computer Science, 9(3–4), 3-4.

Type

Journal article

Intellectual base

Primary research

Year

2014

DOI

10.1561/0400000042

Link

https://doi.org/10.1561/0400000042

Cited from

computer_science, engineering_design, mathematics, neuroscience, statistics_experimental_design

Cited by

8 citations across 8 artifacts.

Each citation links to the sentence it supports in the citing article.

Primes

• Blinding
  • The privacy field generalizes the cut from "remove the information" to "add calibrated noise so the information is recoverable only at known probability"

    This sourceGeneralizes the channel-cut from full removal to adding calibrated noise so a value is recoverable only at a known probability — channel attenuation with a noise wall in place of a clean cut.

• Hidden Information Reconstruction
  • Differential privacy is structurally a channel-narrowing technology; generative-noise training is a prior-denial technology; each addresses a different leg of the reconstruction, and a defence that addresses neither is theatrical.

    This sourceCanonical treatment of differential privacy as calibrated channel-narrowing, the adversary-relative privacy guarantee, and composition theorems quantifying super-additive leakage across disclosures.

• Identifiability
  • And the dual of identifiability is anonymity: the same test with the success criterion flipped, so that differential privacy, k-anonymity, and indistinguishability obfuscation specify minimum sizes of the equivalence class rather than demanding a singleton.

    This sourceTreats privacy as a minimum size of an equivalence class of neighboring databases (the dual of identifiability), with the budget ε sizing the guaranteed indistinguishability.

• Linearity
  • Listed in the references but not attached to a specific claim.
• Measurement Uncertainty and Observational Noise
  • Listed in the references but not attached to a specific claim.
• Observability
  • T3 — Privacy and observability tradeoffs
• Predictive Coding
  • Listed in the references but not attached to a specific claim.

Mechanisms

• Privacy Budget Accounting
  • The discipline is to fix the budget from the protection requirement before demand is known, and to treat an exhausted budget as a real stop rather than a paperwork obstacle.

    This sourceShows how target cumulative-privacy parameters determine per-mechanism privacy allowances under composition.

Verification

This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.

Registry ID ref:80bb106105bd · see in the full table