28.–30. Apr. 2026
DKFZ, Heidelberg
Europe/Berlin Zeitzone

Automated FAIR Evaluation with FAIR-Eva and Its Application to the Coscine Research Data Management Platform

P30
Nicht eingeplant
3m
Communication Center (DKFZ, Heidelberg)

Communication Center

DKFZ, Heidelberg

Im Neuenheimer Feld 280 69120 Heidelberg, germany
Poster 9. From Minimum Requirements to FAIR and AI-Ready: Assessing Metadata Quality POSTERS & DEMOS - with Coffee

Sprecher

David Schimmel (RWTH Aachen University)

Beschreibung

The concept of FAIR Digital Objects (FDOs) supports good research data management practices. However, the evaluation of compliance with the FAIR principles suffers from uncertainties: the FDO specification allows different implementations. Furthermore, heterogeneous, domain-specific terminologies and ontologies are used in concrete FDOs. Access methods for data and metadata differ among FDO implementations. The FAIR principles are open to different interpretations.

Consequently, automatic FAIR evaluation tools often focus on well-established terminologies and FDO providers. In contrast, FIDELIS[1] takes a bottom-up approach with FAIR-Eva[2]. FAIR-Eva is a framework for automatically finding and evaluating metadata properties according to the RDA FAIR maturity indicators. It can guide FAIRification efforts by providing useful feedback to creators of metadata records. A key feature of FAIR-Eva is its interface for defining custom plugins, enabling developers to adapt the tool to their specific metadata schemas.

During a first support round, Coscine[3] was chosen as a service for which plugin development was guided. Coscine is a multi-disciplinary data and metadata management platform. In Coscine, data stewards can register metadata schemas tailored to their domain-specific requirements. Thus, a FAIR evaluation of metadata records cannot completely rely on schema standardization.

Furthermore, the structure of Coscine FDOs is not well-supported by existing FAIR evaluation tools. FDOs are identified by handle-based ePIC PIDs. Each PID is associated with a Coscine-specific Kernel Information Profile, linking to the location of the metadata record. This record describes a complex digital object, comprised of a hierarchy of RDF and non-RDF sources, conforming to the Linked Data Platform. They can be navigated by extensions of the object PID. Automatically following the links and evaluating the correct attributes is a challenge.

We propose presenting FAIR-Eva as well as the challenges and insights from the use case of developing a plugin for Coscine.

[1] https://eosc.eu/horizon-europe-projects/fidelis
[2] https://github.com/EOSC-synergy/FAIR_eva/blob/main/docs/index.md
[3] https://about.coscine.de/

Autoren

David Schimmel (RWTH Aachen University) Herr Fernando Aguilar Gómez (Instituto de Física de Cantabria (CSIC-UC))

Co-Autoren

Ilona Lang (IT Center RWTH Aachen) Marius Politze (IT Center RWTH Aachen) Herr Pablo Orviz Fernández (Instituto de Física de Cantabria (CSIC-UC))

Präsentationsmaterialien

Es gibt derzeit keine Materialien.