This report highlights common challenges and priorities, and proposes a set of initial recommendations on how existing data infrastructures can evolve and collaborate to provide services that support the implementation of the FAIR data principles, in particular in the context of building the European Open Science Cloud (EOSC). The report is an output of three workshops Workshop details: Workshop I, 12 April 2019, Prague (EOSC-hub Week https://www.eosc-hub.eu/events/eosc-hub-week-2019/programme/services-support-fair-data); Workshop II,
24 April 2019, Vienna (Linking Open Science in Austria https://linkingopenscience.univie.ac.at/agenda/); Workshop III, 18 September 2019, Porto (Open Science Fair https://www.opensciencefair.eu/workshops-2019/services-to-support-fair-data-formulating-recommendations-for-eosc) designed to explore, discuss and formulate such recommendations and is aimed at stakeholders in the scholarly world and particularly the EOSC Governance.
This report presents the outcome of an active process of community consultation – most notably in the form of three workshops held in 2019 – to gather, discuss and analyse recommendations for data services and research infrastructures to support the implementation of the FAIR principles. Coming from a broad range of participants, representing several stakeholder groups, these recommendations provide valuable insights into what the participants perceive to be the greatest impediments, challenges, and opportunities for services to support FAIR data. These insights give further direction and impetus to the development of a FAIR data ecosystem as envisioned in the Turning FAIR into reality report, in particular in the context of building the European Open Science Cloud. To deliver tangible and actionable results, with a view of facilitating adoption, the recommendations gathered in the initial two workshops were prioritised and associated with actions and suggested action owners in the third and final workshop. Here it should be clarified that ‘priority’ is meant as a statement of timeliness more than overall value; in other words, participants were explicitly asked to indicate what should be done the most urgently rather than what should be done versus not done.
As introduced and described above, Figure 2 offers a concise summary of the recommendations as well as the relative priority assigned by different stakeholder groups and a panel of experts. As a first observation the picture shows strong heterogeneity, with different stakeholder groups assigning different priorities to the various recommendations (and occasionally disagreeing amongst themselves). This could be a reflection of the relatively low level of maturity with regards to FAIR data, characterised by many simultaneous challenges, limited information or validation of ‘what works’, and various actors reviewing or redefining their roles and responsibilities. Still, an area that seems to stand out and confirmed as a priority is that of essential infrastructure 9 components, including services to automatically create metadata, PID services, and domain-specific ontologies. Complementary to this mostly technical dimension, socially-oriented recommendations around fostering global collaborations and including FAIR in research assessments also scored well across the different stakeholder groups. Suggested actions and action owners for the priority recommendations are collected in Table 1.
The results presented here are naturally a snapshot in time and, as such, represent work in progress. It has proven to be oftentimes challenging to associate more high-level recommendations with pointed, concrete actions and well-defined owners. This workshop series has endeavoured to do that and, while it is hoped that the results presented here will help direct the discussion and spur action, it will no doubt be part of a longer journey with further iterations on the formulation of these recommendations, priorities and actions.
As next steps, the authors were pleased to receive requests to re-use the workshop format to gather and discuss community input in other geographical regions which could help to corroborate findings and paint a fuller, more inclusive picture. In addition, these findings will feed into ongoing work in FAIRsFAIR Notably in Task 2.4 on ‘FAIR services & software’ as well as the development of recommendations in several work packages. See https://www.fairsfair.eu/the-project, the EOSC FAIR Working Group https://www.eoscsecretariat.eu/working-groups/fair-working-group, the Research Data Alliance, OpenAIRE, FREYA https://www.project-freya.eu/en, EOSC-hub Notably in Task 11.2 Data Management Planning. See https://www.eosc-hub.eu/ and other relevant projects. Finally, it is hoped that some readers might recognise themselves as a suggested action owner and find this report helpful to guide them on their path to develop services, infrastructure, tools, ontologies, standards, models, policies and practices that will be supported and valued by the community.
|↑1||Workshop details: Workshop I, 12 April 2019, Prague (EOSC-hub Week https://www.eosc-hub.eu/events/eosc-hub-week-2019/programme/services-support-fair-data); Workshop II,|
24 April 2019, Vienna (Linking Open Science in Austria https://linkingopenscience.univie.ac.at/agenda/); Workshop III, 18 September 2019, Porto (Open Science Fair https://www.opensciencefair.eu/workshops-2019/services-to-support-fair-data-formulating-recommendations-for-eosc)
|↑2||Notably in Task 2.4 on ‘FAIR services & software’ as well as the development of recommendations in several work packages. See https://www.fairsfair.eu/the-project|
|↑5||Notably in Task 11.2 Data Management Planning. See https://www.eosc-hub.eu/|