DPGraph: A Benchmark Platform for Differentially Private Graph Analysis

Differential privacy has become an appealing choice for analyzing sensitive data while offering strong privacy protection, even for complex data types like graphs. Despite a decade of academic efforts in designing differentially private algorithms for graph analysis, few works have been used in practice. This is due to their complexity in the choice of privacy guarantees and parameter/environmental configurations, or due to their scalability issues for large datasets.To bridge the gap between theory and practice, we present DP-Graph, a web-based end-to-end benchmark platform built for re-searchers and practitioners to evaluate private algorithms on graph data. This platform supports a rich set of tunable algorithms for popular graph statistics, such as degree distribution and subgraph counting, with different differential privacy guarantees. A general framework for these algorithms has also been designed for users to tune the algorithms, by changing the sub-algorithms or re-distributing the privacy budget among the sub-algorithms. This enables users to understand the trade-off between privacy, accuracy, and performance of existing work and discover suitable algorithms for their applications.