Search papers, labs, and topics across Lattice.
This paper introduces PPGNN, a personalized differentially private framework designed for decentralized graph data, addressing the limitations of existing methods that apply uniform privacy requirements across users. By implementing a two-stage solution that includes a Personalized Perturbation Mechanism and a weighted calibration strategy, PPGNN allows for user-specific privacy budgets, significantly reducing noise distortion while maintaining analytical utility. Experimental results on six real-world datasets show that PPGNN effectively balances the need for personalized privacy protection with the utility of graph data in learning tasks.
Personalized privacy budgets in decentralized graph learning can significantly enhance data utility while safeguarding sensitive information.
Graph-structured data is increasingly generated and stored in decentralized environments, such as social platforms, mobile applications, and edge networks, where users maintain control over their local graph data. However, collecting and analyzing such decentralized graph data for downstream learning tasks raises significant privacy concerns, as nodes and their attributes often contain sensitive personal information. Local Differential Privacy (LDP) has emerged as a promising solution for privacy-preserving data collection without relying on trusted servers. Nevertheless, existing LDP-based graph learning methods typically assume uniform privacy requirements across users, ignoring the heterogeneous and personalized privacy preferences commonly observed in real-world systems. This uniform treatment leads to inflexible noise injection at the data collection stage, resulting in substantial distortion of graph data and degraded utility in subsequent analysis. To address this limitation, we propose PPGNN, a personalized differentially private framework for decentralized graph data. PPGNN enables user-specific privacy budgets during local perturbation while preserving analytical utility. To handle heterogeneous privacy levels and noise distortion, we design a two-stage solution consisting of a Personalized Perturbation Mechanism (PPM) and a weighted calibration strategy, FlexProp. Extensive experiments on six real-world graph datasets demonstrate that PPGNN effectively balances personalized privacy protection and data utility in decentralized graph learning scenarios.
