TL;DR
This paper introduces a data provenance model specifically designed to help organizations comply with the EU's General Data Protection Regulation (GDPR). By modeling GDPR workflows and providing design patterns, it offers a concrete way to track and demonstrate how personal data is processed, aiding in accountability and verification.
What Problem Does This Paper Address?
The GDPR, in effect since May 2018, imposes strict regulations on how organizations handle the personal data of EU citizens, with significant penalties for non-compliance. A core challenge for organizations is proving that they are indeed complying with these regulations, especially given the global scope of GDPR and the increasing complexity of data processing workflows. Organizations need to demonstrate how data was generated, used, and protected, which requires a detailed understanding of data lineage and processing history. This paper addresses the gap between the legal requirements of GDPR and the technical means to verify compliance, particularly concerning data usage and accountability.
Methods
The authors propose a data provenance model tailored for GDPR compliance. This model is built upon the W3C PROV-DM standard but extends it with specific GDPR concepts. Key aspects of their approach include:
- Explicit Codification: Identifying specific GDPR rights and obligations where data provenance is applicable.
- GDPR-Specific Ontologies: Adapting existing GDPR ontologies to map GDPR concepts to PROV-DM classes and relations.
- Provenance Design Patterns: Developing reusable patterns to represent common GDPR-related events and workflows, such as data collection, consent management, and data transfers. These patterns help in answering compliance questions, enforcing data usage control, and tracing data origins.
- Controller-Centric Model: The model is designed from the perspective of data controllers, who are ultimately responsible for demonstrating compliance.
- Temporal Reasoning: The model incorporates temporal aspects to reason about the validity of consent and data usage over time, which is crucial for GDPR compliance.
Datasets
Not discussed in the paper. The paper focuses on modeling and conceptual design rather than empirical evaluation on specific datasets.
Key Insights and Ideas
- Provenance as a Compliance Bridge: Data provenance is not just a technical concept but a crucial tool for bridging the gap between legal requirements (GDPR) and practical implementation. It provides the necessary audit trail.
- Temporal Nature of Consent: GDPR consent is often time-bound. Provenance is essential for tracking when consent was given, when it expires, and ensuring that data processing aligns with these temporal constraints.
- Accountability Through Traceability: By meticulously tracking data origins, processing steps, and justifications, organizations can establish clear accountability for data handling, which is a cornerstone of GDPR.
- Design Patterns for Practicality: The proposed design patterns offer concrete examples of how to model GDPR scenarios, making the abstract provenance concepts more accessible and actionable for practitioners.
- Privacy of Provenance Itself: Collecting provenance data introduces new privacy considerations. The paper acknowledges that metadata itself needs to be GDPR-compliant, suggesting techniques like hashing personal data.
Key Results
The paper doesn't present quantitative results from an empirical study. Instead, its key contributions are conceptual and methodological:
- A defined GDPR data provenance model that maps GDPR concepts to W3C PROV-DM.
- A set of design patterns illustrating how this model can be used to represent GDPR workflows and answer compliance questions.
- Demonstration of how provenance can be used to verify compliance with specific GDPR requirements, such as tracking data usage after consent withdrawal.
Limitations and Future Work
Limitations:
- Provenance Collection is Not Sufficient Alone: The authors acknowledge that provenance collection and querying are not a complete solution for GDPR compliance. Human oversight and reasoning are still required.
- Annotation Challenges: Annotating existing workflows and applications to collect provenance is often non-trivial and implementation-dependent.
- Privacy of Provenance Data: The collection of provenance data itself raises privacy concerns that need to be addressed, potentially requiring techniques like hashing.
- Standardization Needs: For inter-controller audits, further standardization is needed regarding the granularity of provenance collection, the mechanisms for collection (built-in vs. retrofitted), and tamper-proof solutions.
Future Work:
- Developing practical, implementation-ready provenance collection mechanisms.
- Exploring how to achieve a truly provenance-aware system in practice.
- Addressing the privacy implications of provenance data itself.
- Further research into standardization for inter-controller audits.
Summary updated