DPGraph is a web-based platform to benchmark differentially private algorithms for graph analysis. The goal of DPGraph is to provide an open-source and extendable framework for researchers and practitioners to evaluate existing algorithms for their privacy, accuracy, and efficiency, thus allowing them to discover the state-of-the-art for their applications. This framework also allows researchers to test their new algorithms against existing ones with ease. This repository implements a suite of DP algorithms and contains scripts to evaluate the algorithms over a wide range of graph datasets and privacy budgets.
DPGraph: A Benchmark Platform for Differentially Private Graph Analysis ACM Conference on Management of Data (SIGMOD) 2021
The DPGraph project is lead by Dr. Xi He at the University of Waterloo.
Contributors to this project include:
- Xi He, University of Waterloo, xi.he@uwaterloo.ca
- Karl Knopf, University of Waterloo, kknopf@uwaterloo.ca
- Beizhen Chang, University of Waterloo, beizhen.chang@uwaterloo.ca
- Huiyi Ning, University of Waterloo, hy3ning@uwaterloo.ca
- Sara Qunaibi, University of Waterloo, squnaibi@uwaterloo.ca
- Sreeharsha Udayashankar, University of Waterloo, s2udayas@uwaterloo.ca
- Siyuan Xia, University of Waterloo, siyuan.xia@uwaterloo.ca
- Yihan He, New York University, yihan.he@nyu.edu
- Yuchao Tao, Duke University, yuchao.tao@duke.edu