top of page

Outstanding challenges in federated data reuse

14 JAN 2025


MEETING SUMMARY

The meeting opened by exploring unresolved issues in distributed learning, focusing on Ethical, Legal, and Social (ELSI) challenges tied to adopting FAIR principles, including community resistance and the need for non-technical solutions. The LIFES association was noted to be evolving into a formal structure for sustained collaboration.


Data Governance in Federated Data Reuse - Lauren Maxwell (University Hospital Heidelberg) highlighted the benefits of federated health data reuse and challenges like poor metadata, diverse legal frameworks, and lack of trust. She emphasized improving metadata interoperability, transparency, and detailed descriptions of privacy-utility tradeoffs to address ethical and legal concerns.


Metadata Quality Issues - Priyanka Ojha (Amsterdam UMC) discussed metadata quality challenges in FAIR implementation, including inconsistent terminologies, field requirements, and persistent identifiers. She recommended standardized rules, better validation tools, and prioritizing persistent URLs for better interoperability and integration.


Data Brokering - Jana Miniarikova (Hypherdata) introduced a data brokering platform to streamline cross-border data sharing. The platform simplifies matching data providers with users while addressing regulatory challenges. She called for collaborations to expand features and ensure FAIR alignment.


ODRL for Machine-Actionable Non-Technical Aspects - Rick Overkleeft (4Medbox) presented ODRL (Open Digital Rights Language) as a way to encode permissions, obligations, and prohibitions for machine-actionable data governance. The approach aims to bridge gaps in making ethical and legal constraints interpretable in federated systems.



MEETING NOTES




University Hospital Heidelberg - Federated Data Reuse in Health: Building Trust and Tackling Governance Challenges


Lauren Maxwell from University Hospital Heidelberg highlighted the benefits and challenges of federated data reuse in health, emphasizing the importance of real-time, privacy-preserving approaches to data analysis. She outlined critical obstacles such as low metadata quality, diverse legal frameworks, and lack of trust among stakeholders. Recommendations included to specifically improve metadata interoperability and transparency in modeling approaches to enhance trust and usability in federated systems. Also, to develop comprehensive metadata to describe the technical approach to federated reuse to understand the inherent privacy-utility tradeoffs and ethical or legal concerns for different approaches to federated data reuse.




Amsterdam University Medical Center - Addressing Metadata Quality and Interoperability Challenges in FAIR Data Points

Priyanka Ojha priyankaoe@gmail.com


Priyanka Ojha from the Amsterdam University Medical Center discussed metadata quality issues encountered in implementing FAIR Data Points, citing inconsistent terminologies, mandatory field requirements, and challenges with persistent identifiers. She emphasized the need for standardized rules to enhance interoperability and highlighted gaps in metadata validation processes. Suggestions included better schema validation tools and prioritizing persistent URLs for improved data integration and usage.




Hypherdata - Building a Marketplace for Global Data Sharing

Jana Miniarikova jana@hypherdata.com

Jana Miniarikova from Hypherdata introduced a data brokering platform aimed at simplifying the process of matching data providers with data consumers. The platform facilitates data sharing across borders and systems, addressing logistical and regulatory challenges. She highlighted its potential for automation and called for collaborations to expand its capabilities, ensuring alignment with FAIR principles and user needs.





4MedBox - Leveraging Open Digital Rights Language (ODRL) for Federated Data Reuse

Rick Overkleeft rick@4medbox.eu


Rick Overkleeft from 4Medbox presented ODRL (Open Digital Rights Language) as a framework for addressing non-technical aspects of data reuse in a machine-actionable way. He explained how ODRL can encode permissions, obligations, and prohibitions to improve data governance and trust in federated systems. The discussion emphasized ODRL’s potential to bridge gaps in making ethical and legal constraints interpretable by machines.





Key discussion points

The closing discussion focused on practical solutions to improve compliance with FAIR data principles and federated data reuse. Key challenges included poor metadata quality, lack of trust, and limited accessibility for non-technical audiences. Participants emphasized the importance of rich metadata, clear machine-readable instructions, and user-friendly interfaces to bridge technical and non-technical gaps. Trust was highlighted as crucial, encompassing trust in data, systems, and users.

To make FAIR principles more tangible, participants proposed use cases, such as analyzing the AstraZeneca vaccine withdrawal, to showcase real-world benefits. Cross-disciplinary collaboration was also emphasized, with examples like integrating health and agricultural data systems to address technical and geopolitical challenges. Demonstrations, training programs, and interactive tools were recommended to enhance understanding and adoption.

The session concluded with a call to action for collaboration, including developing digital consent mechanisms, leveraging ethical and legal frameworks, and testing machine-readable templates in federated environments. This multi-faceted approach aims to combine innovation, engagement, and practical demonstrations to advance FAIR principles and federated data sharing across domains.








Comments

bottom of page