Explore the complexities of data anonymization in this 23-minute conference talk from USENIX Enigma 2020. Delve into the challenges of protecting anonymized data as Matt Bishop from the University of California, Davis, examines how adversaries can exploit relationships between data fields to reveal hidden information. Learn about the role of external data in uncovering these relationships and approach data anonymization as a risk management problem. Gain insights into differential privacy, analyze real-world examples like the Netflix and AOL data breaches, and understand the implications for data protection strategies. Discover the importance of considering temporal, situational, and non-obvious relations in data anonymization efforts.
Overview
Syllabus
Intro
The Problem
Key Point to Take Away
The System Model
The Contradiction
Two Goals
Differential Privacy!
About That First One...
Netflix Data
Threat Model
Temporal
Situational
Non-Obvious Relations
Adventures with AOL
First Aftermath
What This Means
Tying Votes to People
Conclusion
A Parting Thought
Contact Information
Taught by
USENIX Enigma Conference
