EDEN: Enforcing Location Privacy through Re-identification Risk Assessment: A Federated Learning ...
summary
EDEN is a device-side system that selects an LPPM and its configuration by running local sanitizations and a federated re-identification assessor called FURIA, avoiding any raw trace upload. It evaluates against a combined attacker ( poi-attack, ap-attack, pit-attack ), measures utility as area coverage, and is validated on five real-world mobility datasets, reporting a consistently better privacy–utility tradeoff than prior centralized or single-LPPM approaches.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:27Motivation
- 0:55Threats
- 1:29Re-Identification Attacks
- 2:12Mobility Profiles
- 2:34Protection Mechanisms
- 3:00LPPM Examples
- 3:38Advanced Systems
- 4:37Problem Illustration
- 4:53EDEN: Overview
- 5:59Results
- 6:35What is next?
- 6:53Reproducibility
related talks
06mPMF: A Privacy-preserving Human Mobility Prediction Framework via Federated Learning
15mA Closer Look: Evaluating Location Privacy Empirically
28mA Privacy Enhancing Federated Learning Framework for Multi-Party Distributed ..
14mEstablishing Privacy Metrics for Genomic Data Analysis
29mBalancing Thin Line Between Data Intelligence & Privacy
46mFrom Health Data to Location Traces: Addressing Spatiotemporal Correlations in Differential Privacy
38mFederated Learning and Its Application...
15mLearning and Unlearning Your Data in Federated Settings