Get 20% off all career paths from fullstack to AI
Master AI and Machine Learning: From Neural Networks to Applications
Overview
AI, Data Science & Cloud Certificates from Google, IBM & Meta — 40% Off
One plan covers every Professional Certificate on Coursera. 40% off Coursera Plus Annual.
Unlock All Certificates
Explore a groundbreaking approach to synthesizing plausible privacy-preserving location traces in this 20-minute IEEE conference talk. Delve into the limitations of existing obfuscation techniques for protecting location privacy and discover a novel generative model that captures both geographic and semantic features of real location traces. Learn how this privacy-preserving framework creates synthetic traces that mimic consistent lifestyles and meaningful mobilities while significantly paralyzing location inference attacks. Examine the statistical similarities between synthetic and real traces, and understand how this method ensures plausible deniability without leaking individual data. Gain insights into the potential applications of this technique in geo-data analysis and location-based services.
Syllabus
Intro
Privacy Preserving Data Publishing
Data Synthesis
Why Synthetic Data?
Location-Based Services (LBS)
Existing Techniques?
Generative Framework
Modeling Human Mobility
Similarity Metrics
Generative Model
Privacy Tests
Utility: What is preserved?
Privacy in LBS Scenario
Conclusions
Taught by
IEEE Symposium on Security and Privacy