A Closer Look: Evaluating Location Privacy Empirically
summary
Empirical evaluation compares local online location-privacy methods on Geolife and Rio bus traces with 300m grid cells and 5‑minute sampling, measuring record metrics (Hamming, Haversine), seven mobility‑pattern features, aggregated queries, and two tasks: co‑location detection (most methods cause high false negatives) and PM2.5 exposure (spatial cloaking gives largest errors). Privacy is tested with re‑identification (most users re‑identified with ≤4 locations) and attribute inference (>90% vulnerable); DP and some traditional methods help, deterministic methods fail under stronger attacks; Java implementation available.
AI-written summary · about
chapters (5)
related talks
01h 20mLocation privacy: what problems - which solutions?
46mFrom Health Data to Location Traces: Addressing Spatiotemporal Correlations in Differential Privacy
23mAnalysis of Privacy Protections in Fitness Tracking Social Networks -or- You can run, but can you hide?
19mDigital Trace Data to Measure Human Mobility
06mWhere You Go Matters: A Study on the Privacy Implications of Continuous Location Tracking
07mEDEN: Enforcing Location Privacy through Re-identification Risk Assessment: A Federated Learning ...
06mThe Crowd Wisdom for Location Privacy of Crowdsensing Photos: Spear or Shield?
06mUnravelling Spatial Privacy Risks of Mobile Mixed Reality Data