🤖 AI Summary
Local differential privacy (LDP) inherently degrades data utility due to mandatory noise injection, and existing post-processing (PP) techniques lack systematic, comparative evaluation. Method: We introduce LDP$^3$, the first comprehensive benchmark platform for PP methods—featuring six LDP protocols, seven PP techniques, four utility metrics, and six real-world datasets, implemented via a modular, multithreaded architecture. Contribution/Results: Experiments demonstrate that PP significantly improves utility under low privacy budgets (ε ≤ 1), yielding up to 35% average gain; however, marginal benefits diminish as ε increases. Crucially, PP effectiveness is highly sensitive to data distribution, utility metric, and analytical task—no universally optimal method exists. This work provides the first quantitative characterization of PP’s applicability boundaries and inherent limitations, establishing a reproducible evaluation framework and empirically grounded guidelines for practical LDP deployment.
📝 Abstract
Local differential privacy (LDP) has recently gained prominence as a powerful paradigm for collecting and analyzing sensitive data from users' devices. However, the inherent perturbation added by LDP protocols reduces the utility of the collected data. To mitigate this issue, several post-processing (PP) methods have been developed. Yet, the comparative performance of PP methods under diverse settings remains underexplored. In this paper, we present an extensive benchmark comprising 6 popular LDP protocols, 7 PP methods, 4 utility metrics, and 6 datasets to evaluate the behaviors and optimality of PP methods under diverse conditions. Through extensive experiments, we show that while PP can substantially improve utility when the privacy budget is small (i.e., strict privacy), its benefit diminishes as the privacy budget grows. Moreover, our findings reveal that the optimal PP method depends on multiple factors, including the choice of LDP protocol, privacy budget, data characteristics (such as distribution and domain size), and the specific utility metric. To advance research in this area and assist practitioners in identifying the most suitable PP method for their setting, we introduce LDP$^3$, an open-source benchmark platform. LDP$^3$ contains all methods used in our experimental analysis, and it is designed in a modular, extensible, and multi-threaded way for future use and development.