A hybrid anonymization pipeline to improve the privacy-utility balance in sensitive datasets for ML purposes.
Jenno VerdonckKevin De BoeckMichiel WillocxJorn LaponVincent NaessensPublished in: ARES (2023)
Keyphrases
- differential privacy
- privacy preserving
- privacy guarantees
- anonymized data
- privacy requirements
- data anonymization
- information loss
- privacy protection
- social network data
- personal information
- privacy preservation
- preserving privacy
- data publishing
- maximum likelihood
- data privacy
- private information
- privacy preserving data publishing
- information systems
- utility function