Takeaways from the EOSC-Nordic WP4 Workshop “From Self-Assessment to Certification with FAIR Results”
On June 3, EOSC-Nordic WP4 hosted a virtual workshop for Nordic and Baltic data repositories currently undergoing or interested in repository certification. While primarily aimed at participants engaged in certification assessments, the event was open to all – and attracted approximately 40 registrations.
The event was intended to be interactive, allowing participating repositories to share experiences, and ask questions while learning more about how EOSC-Nordic and FAIRsFAIR interact and how this could help repositories improve their data management processes and policies and increase their trustworthiness. Nearly 40 minutes had been reserved for discussion, and it was used until the last minute!
Summary of the topics presented
In the introduction, Mari Kleemola from the Finnish Social Science Data Archive noted that the main goal of EOSC-Nordic WP4 is to build up and facilitate the practice of depositing FAIR digital objects to trustworthy digital repositories. This view is also expressed in the Turning FAIR into Reality report by the European Commission Expert Group.
Linas Čepinskas from DANS (Data Archiving and Networked Services) presented an overview of the best practices identified by FAIRsFAIR. The main goal is to bring practical solutions to the implementation of FAIR principles throughout the research cycle. Several project partners provide certification support for research data repositories. Two tools have been released to evaluate FAIRness: Fairaware, which helps the researcher to assess the FAIRness of a data publication, and F-UJI to assess the FAIRness and machine-actionability of a research data object.
EOSC-Nordic WP4 takes advantage of the F-UJI tool in FAIR maturity evaluation, and this is not the only form of collaboration between projects. Geographically, FAIRsFAIR operates mainly in Western / Southern Europe and EOSC-Nordic in the Nordic and Baltic countries.
Philipp Conzett from the UiT, the Arctic University of Norway, shed light on the CTS certification process. DataverseNO, a generic repository for open research, was certified in 2020. The decision to apply for certification was made to express commitment to FAIR data stewardship. Philipp noted that certifying an entire distributed repository was a complex task. One of the major difficulties was establishing a full-scale preservation plan. To address this, they created separate plans for different asset groups. They noticed, however, finding good examples to learn from proved to be a challenge. In the ensuing discussion, Linas pointed out that FAIRsFAIR had adapted the DPC Policy Toolkit elements to their Preservation Policy Planning Worksheet.
Kari Lahti from FinBIF, the Finnish Biodiversity Information Facility, provided an overview of why FinBIF is applying for certification. Certification means actively building trust, assessing and learning about your performance, increasing your potential to attract more funding, and becoming a valued partner. Kari pointed out that the journey is the most critical stage, not only the goal. If support is available, it is an opportunity not to be missed. In any case, it pays off to review the certification requirements, approaching them step-by-step. They are not always self-evident, and terminology may not be familiar. Certification is likely to be time-consuming if one lacks good documentation of the repository’s mission and processes.
Synergies of FAIR and CTS
FAIRsFAIR is mapping the FAIR principles to CoreTrustSeal requirements, and WP4 will evaluate how CTS+FAIR certification might look in practice. Although FAIR principles and repository certification are not designed to achieve the same goal, there is much overlap in practical implementation solutions.
Currently, there is very little focus on machine-actionability in CTS. It is essentially a human-readable evaluation while the FAIR principles were designed for machines. However, investing in both support each other in many ways. Together they help the community to assess the scholarly quality of the data. Lots of research data is available, and the human brain cannot process it all. Machine-actionable (meta)data is needed, but researchers, data curators, and managers are needed to produce the metadata.
The FAIR principles do not in themselves imply any particular form of implementation. Go FAIR provides some examples of real-world implementations. However, the scholarly communities are free to put the FAIR principles into practice as they see fit.
FAIR is not the same as “open” and “free,” and CoreTrustSeal may be granted to a data repository that makes data reuse possible – openly or only accessible under predefined conditions. Terms may vary and are not limited to commonly used Creative Commons licenses. Therefore, both criteria can be applied to highly sensitive data, such as health data, where strict access control practices are required.
Recommendations for future work are due this year. The latest FAIRsFAIR outputs are available on Zenodo and EOSC-Nordic outputs on the project’s Knowledge Hub.
Webinar presentations are available on the EOSC-Nordic Knowledge Hub Training Library.
Welcome and Introduction to EOSC WP4 Goals and Support Modes Available
Mari Kleemola (Finnish Social Science Data Archive)
Making your repository more trustworthy – good practices from FAIRsFAIR
Linas Čepinskas (DANS – Data archiving and Networked Services)
Repository Experiences on Certification: DataverseNO
Philipp Conzett (UiT The Arctic University of Norway)
Repository Experiences on Certification: FinBIF
Kari Lahti (FinBIF – Finnish Biodiversity Information Facilit
Author: Tuomas J. Alaterä
Finnish Social Science Data Archive, EOSC-Nordic WP4 member