DPGen: Automated Program Synthesis for Differential Privacy

In this paper, we present DPGen, a program synthesizer that takes in non-private code (without any noise) and automatically synthesizes its differentially private version (with carefully calibrated noise). Under the hood, DPGen uses novel algorithms to automatically generate a sketch program with candidate locations for noise, and then optimize privacy proof and noise scales simultaneously on the sketch program. Moreover, DPGen can synthesize sophisticated mechanisms that adaptively process queries until a specified privacy budget is exhausted. When evaluated on standard benchmarks, DPGen is able to generate differentially private mechanisms that optimize simple utility functions within 120 seconds. It is also powerful enough to synthesize adaptive privacy mechanisms.

Files

Metadata

Work Title DPGen: Automated Program Synthesis for Differential Privacy
Subtitle Proceedings of ACM Conference on Computer and Communications Security (CCS)
Access
Open Access
Creators
  1. Wang Yuxin
  2. Zeyu Ding
  3. Yingtai Xiao
  4. D Kifer
  5. Danfeng Zhang
License In Copyright (Rights Reserved)
Work Type Article
Publication Date January 1, 2021
Deposited September 27, 2022

Versions

Analytics

Collections

This resource is currently not in any collection.

Work History

Version 1
published

  • Created
  • Added ccs21.pdf
  • Added Creator Wang Yuxin
  • Added Creator Zeyu Ding
  • Added Creator Yingtai Xiao
  • Added Creator D Kifer
  • Added Creator Danfeng Zhang
  • Published
  • Updated