Research Data Management: where are libraries now?

Are libraries the right place to manage Open research data and, if they are, how much progress have we made towards doing it?  This was a major theme of the 2013 LIBER conference, explored in at least seven presentations and workshops.  The consensus was largely that libraries had a significant role to play, and a number of institutions have started to explore the boundaries of what is possible within libraries – advocacy for Open data, support for metadata creation by researchers, helping to write data management plans, and curating data output from “small science”.  LERU (League of European Research Universities) is developing a roadmap for research data which sets out this list of roles for libraries.  Paul Ayris outlined their work in his presentation and went on to elaborate its implementation at UCL.

Top concerns for libraries included:

  • Research council requirements: What research data management services need to be provided to comply?
  • Infrastructure: How do we make the business case for investment?
  • What roles and responsibilities are right for libraries?
  • Are we recruiting and training the right staff? Are librarians capable of developing the technical skills required?
  • If we recruit or train data librarians, how do they gain an understanding of research practice?  Should they be embedded in research groups (and possibly become isolated) or the library?  

Liz Lyon (UKOLN and DCC), in her plenary presentation, Roadmaps, Roles and Re-engineering: Developing Data Informatics Capability in Libraries, provided the clearest guidance, as always, on where to start.  Looking at the implications of the EPSRC policy on research data as an exemplar, she outlined the requirements, and some models and tools that we might might use: the example of the business case written for her own university (Bath), the DCC simplified data management plans checklist, and Bath’s own DMP guidance and template.

She went on to look specifically at institutional data publication services, which could be library based – data curation in the repository, cataloguing and discovery, citation, and metrics – and source material for skills development.  Data librarians are few and far between but training opportunities are beginning to develop.  Liz Lyon highlighted the new ImmersiveInformatics pilot course at the University of Bath, co-developed with University of Melbourne.

Geoffrey Boulton (University of Edinburgh), author of the Royal Society’s policy paper on Science as an Open Enterprise, outlined in his presentation, A Revolution in Open Science: Open Data and the Role of Libraries, the benefits of open data and sharing (identifying fraud in science, public accountability, economic benefits, rapid response to medical emergencies, crowd-sourcing, citizen science), but stressed that the data needs to be intelligible. He was slightly sceptical about whether libraries could take on the challenge, perhaps provoking us to start building it into our strategies.  Universities need to take on responsibility for the data they produce but have to be proactive, not just compliant.  They need strategies, e.g. in the library, and management processes.

Funding the infrastructure and services and making them interoperable if data management responsibility lies with multiple organisations, remains the most significant issue.  Carlos Morais Pires (DG Connect, European Commission), Enabling Data-Intensive Science through Advanced Data e-Infrastructures and Services, spoke on the EC’s Horizon 2020 programme and consultation, which was an outcome of the Riding the Wave report.   Its vision was “data e-infrastructure that supports seamless access, use, re-use, and trust of data. In a sense, the physical and technical infrastructure becomes invisible and the data themselves become the infrastructure a valuable asset on which science, technology, the economy and society can advance”.

The implemention of an interoperable data infrastructure is envisaged as:

(a) data generators; research projects, big research infrastructure, installations or medium size laboratories, simulation centres, surveys or individual researchers
(b) discipline-specific data service providers, providing data and workflows as a service
(c) providers of generic common data services (computing centres, libraries)
(d) researchers as users, using the data for science and engineering

There will be funding opportunities in the Horizon2020 programme.  It is currently out for consultation and has received responses from research centres (e.g. CERN) individual universities, and libraries organisations, including LIBER, and the League of European Research Universities (LERU).

Further RDM-related presentations explored cost modelling storage of Print and Digital Collections (Darryl K. Mead, National Library of Scotland), a European policy framework for open access to research data (Susan Reilly, LIBER), initiatives to upskill IT and research support staff for RDM in France (Marie-Christine Jacquemot-Perbal, Inist-CNRS, and collaborative RDM projects in the MLA sector in Denmark.

All presentations can be found on the LIBER 2013 web site.