ARX Data Anonymization Tool

ARX Data Anonymization Tool provides anonymization of sensitive personal data to enable analysis and sharing while protecting privacy in clinical, epidemiological, and research datasets.

Key Features:

• Privacy and Risk Models: Supports diverse privacy models and risk assessments to tailor anonymization and address singling out, inference, and linkage risks.
• Data Transformation Methods: Implements generalization, suppression, perturbation, and synthesis to reduce re-identification risk while preserving analytical value.
• Usefulness Analysis: Computes measures of data utility to evaluate the impact of anonymization on data quality.
• Scalability and Performance: Processes large datasets efficiently on commodity hardware to support high-volume data anonymization.

Scientific Applications:

• Epidemiological registry anonymization (LEOSS, SARS-CoV-2/COVID-19): Applied quantitative anonymization procedures to the Lean European Open Survey on SARS-CoV-2 Infected Patients (LEOSS) to enable public real-time sharing of COVID-19 patient registry data.
• Clinical trial data sharing: Enables anonymization of clinical trial datasets to permit secondary analysis while protecting participant privacy.
• Big data analytics and research projects: Supports anonymization workflows for commercial big data analytics platforms and academic research datasets.
• Educational and training datasets: Facilitates creation of anonymized datasets for educational and training purposes without exposing sensitive personal information.

Methodology:

Applies privacy models and risk assessments, uses generalization, suppression, perturbation, and synthesis, performs usefulness analysis, and employs quantitative anonymization procedures to protect against singling out, inference, and linkage attacks; research reports minimal introduced bias.