PRIVEE: A Visual Analytic Workflow for Proactive Privacy Risk Inspection of Open Data.

Area of Science:

• Computer Science
• Data Privacy
• Information Security

Background:

• Open data sets, even when anonymized, are vulnerable to privacy breaches via data joins.
• The release-and-forget model of data publication neglects potential privacy risks.
• Malicious actors can exploit shared attributes across datasets to compromise individual privacy.

Purpose of the Study:

• To address the gap in proactive privacy risk assessment for open data.
• To develop a visual analytic solution for data defenders to identify disclosure risks.
• To enable data custodians to understand and mitigate privacy vulnerabilities in joinable data.

Main Methods:

• Conducted a design study with data privacy researchers, involving ethical hacking and privacy attack simulations.
• Developed visual analytic interventions based on identified attack scenarios.
• Realized these interventions in PRIVEE, a visual risk inspection workflow.

Main Results:

• PRIVEE provides data defenders with awareness of disclosure risks in local, joinable data neighborhoods.
• The solution uses risk scores and interactive visualizations to explore vulnerable joins.
• It allows interpretation of risks at multiple data granularity levels.

Conclusions:

• PRIVEE acts as a proactive monitor for data defenders against privacy attacks.
• The system helps emulate attack strategies to diagnose disclosure risks effectively.
• Case studies with experts demonstrate PRIVEE's utility in enhancing data privacy awareness and defense.