Five workshops were co-organised by FAIRsFAIR, OpenAIRE, FIT4RRI and EOSCSecretariat.eu during OS Fair 2019 in Porto, and united a variety of user perspectives around aspects of Open Science, scholarly communication in national settings, and FAIR implementation. The key learning points and outcomes are summarised here. Links in each workshop title are to the original agenda.
17 September, 11:00
The European Commission FAIR Data Expert Group report Turning FAIR into Reality prioritises “the increased provision and professionalisation of data stewardship”, and states that “New job profiles need to be defined and education programmes put in place to train the large cohort of data scientists and data stewards required to support the transition to FAIR.”
Incorporating a mix of case study presentations and group discussions between practitioners, policy makers, and service providers, this workshop explored the challenges inherent in engaging, recognising and retaining data stewards, and evaluated progress to date.
Amongst those present, the main challenges to professionalisation include educating management, hiring, training, and mapping out a career path for personnel, and building an international data stewardship community.
Skill and training frameworks and job profiles exist or are in development. Examples:
- FAIR4S Skills Framework defined under the EOSCpilot WP7 on Skills and Capability.
- Data Stewardship on the Map study undertaken by the Netherlands National Coordination Point on Data Management (LCDRM) in which a task group of more than 30 institutions worked to define clearer job descriptions for data stewards.
- DTL (Dutch Tech Centre for Life Sciences) project Stewardship in the Life Sciences. See presentation by Celia van Gelder, Programme Manager DTL Learning/ELIXIR-NL Training at https://zenodo.org/record/3420179#.XYziI25uKM9.
- The Research Software Engineering (RSE) community which offers a blueprint for the professionalisation of data stewardship. The organisation has established defined roles, formed an association, and holds regular conferences.
Aligning the work and outputs of the different initiatives underway is a priority. Actions going forward include setting up a European interest group and creating a common definition of data stewardship and a centralised repository of training materials and resources. Producing a trainer handbook is also proposed.
17 September, 14:00
The focus of this workshop was sharing success stories of FAIR data generation and exploring ways to extend this success to other domains.
Amongst those who presented case studies were ambassadors from EOSC-Pillar, EOSC-Life, FAIR Plus, PANOSC, FAIR4HEALTH, the Libraries for Research Data Interest Group, and the EOSC Secretariat Executive Board Working Group.
Participants agreed on the importance of defining a policy framework and standardising tools and norms. However, given the necessity for data to be managed according to the needs of practitioners, adopting common FAIR methodologies starts currently in individual disciplines - typically within thematic ESFRI cluster projects, and is supported either transversally by multidisciplinary initiatives such as FAIRsFAIR and the Libraries for Research Data Interest Group, or thematically, by FAIR4HEALTH, the FAIRplus project, and others.
For example, in the case of PANOSC, which contributes to the realisation of a data commons for neutron and photon science, guidelines on bringing FAIR to the proceedings are provided by ExPaNDS (EOSC Photon and Neutron Data Services). For life sciences, the FAIRplus project will launch the FAIR Cookbook (beta version December 2019) with “recipes” on how to make data FAIR and support the data management culture in pharma, academia, and amongst SMEs.
From an interdisciplinary perspective, the EOSC Secretariat Executive Board Working Group is one source of recommendations on FAIR implementation. Deliverables include metrics for FAIR data and repository certification, a persistent identifier policy, and an EOSC interoperability framework by using fair practices. Fairsharing.org is a curated resource on data and metadata standards, and helps producers to make their resources more visible, and more widely adopted and cited.
17 September, 14:00
The purpose of this workshop was to review current practice around publishing and sharing EOSC training resources, events and tools, and identify metadata and other features suitable for adoption into a common framework for a sustainable training structure.
Participants included representatives from the EOSC Working Group on Training and Skills, EOSC Portal, and LIBER as well as EOSC Pilot, the project tasked with developing the FAIR4S common standards and assessment framework of FAIR data stewardship skills for science and scholarship, and draft recommendations on FAIR training.
- Rather than reinventing the wheel, the intention is to assess existing initiatives, inventory best practice and work towards adoption at national level.
- As such, collaboration between regional and national initiatives and coordination of activities is fundamental.
- Certification of training materials and trainers is recommended.
- The competence centre concept is not well defined amongst audiences and should be clearly defined.
- Researcher needs should be more clearly identified and addressed.
- Documenting courses with metadata schema is essential for the FAIRification of training courses.
- A methodology should be found for keeping the list of services up to date in the EOSC portal.
17th September, 14:00
Persistent identifiers (PIDs) are at the heart of the FAIR principles and are becoming increasingly important in all research workflows. This workshop explored the identifier landscape and examined the use of identifiers in disambiguating and linking research.
- PIDs are becoming increasingly important for research
- While individual PIDs add value, the true power of PIDs can be seen when PIDs and PID systems are connected.
- Different stakeholder groups are working on PIDs and their approaches and efforts converge. Because the space is complex, clear policies and training materials are important..
- Researcher engagement often focuses on ORCIDs and less on other identifiers. Whilst engaging researchers around PIDs is a challenge, the topic of PIDs should be pursued as a contributor to the FAIRness of data. All processes should be rendered as seamless as possible.
- Although PIDs are intuitively valuable, it would be helpful to have more direct evidence, for example, metrics, showing the value of PIDs.
18 September, 09:00
This was the last event in the Services to Support FAIR Data series. The aim of this workshop was to review previously tabled recommendations with an expanded group including providers and users of e-infrastructures, national representatives and communications experts from academia, and then to move beyond recommendations and formulate some clear actions along with a view on who should be taking these actions forward.
A detailed analysis of the outcomes of this particular workshop has been published and is available for review here.