Probabilistic Fingerprinting Scheme for Correlated Data

Area of Science:

• Computer Science
• Data Security
• Cryptography

Background:

• Individuals share personal data for services, risking confidentiality.
• Unauthorized data distribution by service providers necessitates source identification.
• Existing fingerprinting schemes are vulnerable to correlation-based attacks.

Purpose of the Study:

• To propose a robust and efficient probabilistic fingerprinting scheme for personal data.
• To address vulnerabilities in existing schemes against data correlation attacks.
• To ensure accountability for unauthorized personal data sharing.

Main Methods:

• Developed a probabilistic fingerprinting scheme considering data utility and inherent correlations.
• Integrated Boneh-Shaw fingerprinting codes for enhanced robustness against collusion.
• Implemented and evaluated the scheme on real genomic data.

Main Results:

• The proposed scheme efficiently generates fingerprints while maintaining high data utility.
• Demonstrated robustness against collusion using Boneh-Shaw codes.
• Experimental results confirm the scheme's efficiency and robustness on genomic data.

Conclusions:

• The probabilistic fingerprinting scheme effectively traces data leaks from service providers.
• The scheme provides a robust solution for data owner accountability.
• This approach enhances personal data security in the context of service sharing.