A National Research Data Service for the UK?
UK Research Data Service Feasibility Study
Stephen Pinfield
The potential
“Because digital data are so easily shared and replicated and so recombinable, they present tremendous reuse opportunities, accelerating investigations already under way and taking advantage of past investments in science.”
Clifford Lynch, Executive Director of CNI, ‘Big data: How do your data grow?’ Nature 4 September 2008
HEFCE Shared Services programme
• HEFCE call for bids in spring/summer 2007• RUGIT and CURL (now RLUK) submitted bid• Both groups concerned about “data deluge”• No UK-wide approach yet visible• HEFCE approved the bid and ITT was drawn up for
major consultancy autumn 2007• JISC also contributed funding
The approach
• SERCO appointed as consultants February 2008• Project Manager appointed in March• Governance from Steering Committee and Project
Management Board• Over 40 stakeholder bodies identified • Iterative process of communication via mailing lists
and personal briefings
What’s the problem?
• “Data deluge”• Research data – an untapped resource• Often unstructured and inaccessible• Lack of coherent policies and standards• Many formats and disciplines• HE Library and IT services under pressure to help
(mainly storage) - unsustainable
What’s the problem?
• Whole data lifecycle, not just storage
• Creation, selection, ingest, storage,
metadata, retrieval, review, preservation
• Access/analyse/synthesise others’ data
• It is the management of the data that needs a UK-wide approach
Aims of the project
• Develop an understanding of the UK’s current and future research data service needs
• Identify gaps in current services
• Test the feasibility of a UK-wide shared service for the management of research data
• Avoid reinventing the wheel in any proposed solution
Methodology – case studies
• Four case study universities: Bristol, Leeds, Leicester, Oxford (April to June 2008)
• Questionnaires and focus groups at first three
• Complementary internal project at Oxford dovetailing with UKRDS
• Total number of individuals consulted: 700+
What did we learn?
• Over 360% growth in data volume expected over the next 3 years
• c50% of data estimated to have a useful life of up to 10 years; 26% seen as having indefinite retention value
• Most research data is currently held locally
What did we learn?
• 21% use a national or international facility
• Most share data within research teams (12% do not); 18% share via a data centre
• 43% would like access to others’ data
• Those with no access to a national facility are particularly keen on a UKRDS
Methodology – desk research
• Ongoing throughout the project
• Finding out what services already exist in UK
• Speaking to key service providers
• Following initiatives in other countries, notably USA, Canada, Australia, Europe
• Fast developing topic
What are we learning?
• National data centres with considerable skills and resources which could be spread
• Digital Curation Centre Life Cycle Model• Data management plans (Wellcome)• JISC’s integrated infrastructure and JANET• JISC and RIN studies (eg data handling skills,
preservation costs)
What are we learning?
• The thesis posited in the bid to HEFCE is borne out by the case studies and desk research
• Research data are important and there is a gap to be filled
• A coherent UK-wide approach involving existing services is feasible
Interim report recommendations
• UKRDS should add coherence to what is already there and:• Be governed by a set of policies• Deliver a set of processes• Support a set of standards• Be measured by an agreed set of KPIs• Be funded to ensure it can be relied on long term• Steering Committee approved a “hybrid” model as opposed to
“do nothing” and “create a new central body” options
Venture Capitalists
Commercial users & generators
of data
Public sector users & generators
of data
Other educational institutions
Vendors
UKRDS
Service providers
International links
Journal and data publishers
HEIs & Research Institutes
• Researchers• IT Directors• Librarians• Archivists
• other experts
Facilitate provision of persistent citation links
Engage as appropriate to maximise exploitation of financial
support for long-term data management capabilities
Funders
Work with funders on policy issues and data management planning
Ensure provision of accession and
access procedures
Coordinate capacity planning and help
address implications for long-term storage and
infrastructure investment
Provision of conditional
data set access
Services covering: data management advice, DCC lifecycle adoption and guidance, training in DMPs, tools /
discovery development, and accession planning
Engage as appropriate to maximise exploitation of
vendor support for long-term data management capabilities
Communities and headline processes
Communities
Processes defined by needs of the Communities
Venture Capitalists
Commercial users & generators of
data
Public sector users & generators
of data
Other educational institutions
Communities of interest
EarlyAdopters
LaterAdopters
Vendors .
UKRDS
Funders
Service providers
International links
Journal and data publishers
HEIs & Research Institutes (academic
status)
• Researchers• IT Directors• Librarians• Archivists• other experts Others
.
Are we in step?
• US model, distributed and NSF-funded: 5 large Datanets (consortia of universities) to build data stewardship capabilities - $100m over five years
• Australian model, centralised and top-down from government: ANDS (Australian National Data Service) with AUS$ 24m over 3 years
Are we in step?• ANDS – “funded to start the process of helping
researchers deal with this deluge” by:• “addressing issues of research data ownership and
responsibilities “• “providing access to research data collected and
maintained with public funding”• “co-ordinating and sharing best practice for curation
of experimental, research and published data”
Are we in step?
• Research Data Canada; government agency says there is “urgent need for action to propel Canada into a new and transformational data-intensive paradigm for Canadian research” and:
• “The research process generates huge amounts of data that are an important part of Canada’s scholarly record and hold enormous potential as an additional discovery and problem-solving tool for researchers”
Are we in step?• German science organisations alliance wants to
equip scientists/scholars “with the infrastructure best suited to meeting their research needs” and states:
• “In the digital age, this entails digital access to publications [and] primary research data, available to the user without costs or other barriers”
• This is in the explicit context of “increasing Germany’s competitiveness in world-class research”
Current work on UKRDS
• Estimate likely costs of filling gaps versus costs of developing provision without a UKRDS
• Develop a value proposition and process models• Design service portfolio and outline governance
model• Write business case and funding proposition• Produce final report for HEFCE December 2008
The next steps
• Seek interim HEFCE/JISC funding in 2009 for a ‘Pathfinder’ service in co-operation with case study institutions and some existing providers
• Hold an international conference 26 February 2009 to promote the proposed service and highlight developments in the UK and abroad
Key messages
• The study is looking for the optimum business case to address the sustainability of what researchers need
• It is not just about storage
• It is about the management of the whole data lifecycle
• It will not reinvent wheels
Key messages
• Many building blocks are already in place
• A UKRDS would embrace rather than replace existing facilities, providing a framework
• There are also significant gaps to be filled
• It is about the leverage of more research value and a higher global research reputation for the UK
Conclusion
• A UKRDS is likely to be feasible
• Need to work out a practical business case and organisation to make it happen
• Need support of funders, HEIs and other stakeholders to get it established and stable for the long term
