1. Case study 2: Research Data
Saskia Franken, Utrecht University
LERU workshop, Brussels
June 6 2014
This presentation is licensed under a Creative Commons Licence
Attribution-NonCommercial 3.0 Unported (CC BY-NC 3.0)
3. Utrecht University
since 1636
• All disciplines
(except Agriculture & Technical Studies)
• 7 Faculties
• Two campuses
• 30.000 students
• 8,500 staff
• Budget € 730.000.000
Utrecht University:Utrecht University:
Curiosity-driven,Curiosity-driven,
relevant to societyrelevant to society
4. Data are hot …
• Recent cases of scientific fraud > discussion about integrity
• Commissie Schuyt in the Netherlands: a plea for openness
of research data
• NWO (main Dutch funder) policy: data that arise from
NWO-funded research should be made publicly available
• EU policy
5. Already in 2010 at Utrecht University
Library
• 2010 start of a pilot research data, resulting in
Utrecht Dataverse Network
• Why?
- Utrecht University Library is Partner in Science
- Utrecht offered already “open” services in storing and
presenting publications (repository)
- Questions of researchers and research projects who were
looking for a safe and easy tool to store their research
data, reached the library
6. 2010 - 2014 Utrecht Dataverse Network
developed into Dutch Dataverse Network
• ,
7. 2014: transfer of Dataverse to DANS
• DANS is backoffice. Tasks: maintenance, upgrades, new
developments, helpdesk
• University libraries are frontoffice. Tasks: contact with the
researchers
Organisation
• Regularly meetings between DANS and libraries about
functionality
• Libraries are in an Advisory board at DANS
16-03-15
9. 2013: Utrecht University starts with a
Research Data policy framework
• Inspired by the LERU roadmap for Research Data
10. Making a research Data policy framework
• Joint venture of policy department, ICT department and
library
LERU roadmap says:
“Research Data Policy, Technology and Support: to promote
succesfull use of research data, these three aspects should
be offered simultaneously to researchers. …. Coordinated
and parallel approach is therefore crucial.”
11. Start: consultation of the 7 faculties
• Utrecht University doesn’t works top-down, especially not
in the field of research
• Faculties are quite autonomous in research matters
LERU roadmap :
“ Institutions and stakeholders engagement. Many
institutional stakeholders are involved in the research data
lifecycle..”
12. Results of the consultation
• Big differences between faculties and within faculties (and
disciplines)
• Some faculties are quite ahead, in developing a policy
(Faculty of Social and Behavourial Sciences) or in
infrastructure (Faculty of Geosciences, Medical Sciences)
• But all are more or less in need of a university policy
framework / a code of conduct. (Not too rigid!)
• Other issues: infrastructure, tools, (longterm) maintenance
of research data, costs, sharing of data.
13. 2014: making an university policy
framework. Draft is ready now.
LERU roadmap says:
“Each LERU member should consider developing an
institutional roadmap.”
“Each LERU member should develop and promulgate an
institutional data policy which clarifies institutional roles and
responsibilities for RDM”
14. Content overview
•Definition of research data
•Goal and scope of our policy
•Regulations and rules that lay on the basis of our policy (f.i.
national laws, internal regulations)
•General conditions that must meet RDM at Utrecht University
•Roles and responsibilities
•What’s next: Utrecht University’s roadmap
16. Roles and responsibilities
•UU has made a precise description of roles and
responsibilities, subdivided in
1. Researchers and (PHD) students
2. Managers of research departments
3. Dean of the faculty
4. University Board of Directors
LERU roadmap says:
“To outline roles and responsibilities is the first step for research
institutions. ”
17. Data management plans
•Researchers are required to deliver a Data Management Plan
at the start of their project and execute the agreements
made in the plan during their project.
LERU roadmap says:
“Research funders increasingly require data management plans… The
LERU research community needs to take note on this development”
18. Storage and archiving
•Research data have to be safely stored and archived for at
least 5 years
Data curation and preservation gets a lot of attention in the LERU
roadmap.
19. Accessibility and sharing
• Research data should be made available for access and re-
use for scientific research inside and outside Utrecht
University “if reasonably possible” and “subject to
appropriate precautions”
LERU roadmap says:
“… the growth of open data must be underpinned by a shift towards a
culture of open access”
“Data should me made open at the most appropriate time. Not all
data can be open”
20. Roadmap
•Next steps, after the policy document is settled:
1. Advocacy
2. Training (f.i. data management courses PHD students)
3. Infrastructure
4. Support
LERU roadmap says:
“LERU universities should establish an asset of research data
facilities and a portfolio of tools”
“LERU universities should organize … a coherent support service ”
21. Meanwhile
•EUDAT pilot
•Developing an Utrecht University DMP template (online)
•Further investigation needs and wishes of researchers
ICT department and library working close together in this
23. • Dutch initiative (10 institutions) to integrate EUDAT
services with universities’ data infrastructures
• Aim: use and build upon the EUDAT services
– to facilitate researchers
– to manage research data and
– to collaborate on research data
– across universities/research institutes
U2CONNECT
Integrity also an important issue in strategic plan UU.
In the Utrecht Dataverse Network (DVN)
researchers can store their data in an online environment,
index these data in a user-friendly way
and share the data with other scientists. The researchers decide who
gets access to what data.
DANS doet back office: beheer, upgrades, ontwikkeling functionaliteit en helpdesk. Universiteiten doen front office: ondersteuning van onze eigen onderzoekers. Contract met universiteiten via bibliotheken. Bibliotheken in a) beheeroverleg en b) adviesraad bij DANS. 7 instellingen zijn overgegaan van ons naar DANS (als het goed is, is iedereen mee, ik heb daar geen bewijs van, maar denk dat wel) per 1 mei. Goede aansluiting op EASY van DANS zodat data permanent bewaard kunnen worden. overdracht gedaan omdat het te groot werd voor ons om te beheren: als bibliotheek is je core business de universitaire gemeenschap en niet collega’s als klant.
Goal: offer a framework so that quality, availability and accessibility of research data is secured and describe what requirements Utrecht University proposes to good RDM.
Scope: it’s a framework, no detailed policy document. No mandate, no obligations. Research at UU is too diverse to proclaim institutional rules. Layered model. Delegated/shared responsibilities: with faculty / research department and researcher
Conditions that must meet good RDM: storage, curation accessibility etc.
Also activities in all other Dutch universities. Examples: Amsterdam (policy). Delft (support), Wageningen (DMP for PHD students obligatory)
EUDAT is an European intiative, financed by FP7, meant to make available generic data services. EUDAT is a consortium of 26 full and 7 associated partners, including eight research communities coming from different research areas, zoals CLARIN (Linguistics), diXa (Chemical Safety) EPOS (Seismology, Volcanology, etc) LifeWatch (Biodiversity) etc.
B2Share: Een repository service waar data gearchiveerd kan worden en een ‘persistent identifier’ krijgt. Het is erg laagdrempelig wat een voordeel is, maar het nadeel is dat er maar erg weinig metadata gevraagd wordt en dus vindbaarheid mogelijk later een probleem wordt.B2Find: Een portal waar ook data uit andere community repositories gezocht kan worden. Op het moment zijn dit DataCite, Clarin, B2Share, NARCIS, GBIF, ENES, SDL. Dit is natuurlijk erg handig, maar een probleem op het moment is dat de community repositories niet alles beschikbaar stellen voor harvesting door B2Find.B2Stage: Inname en vervoer van grote hoeveelheden data tussen EUdat opslag plekken en High Performance Computing (HPC) systemen. Dat kan voor sommige onderzoekers erg aantrekkelijk en zelfs noodzakelijk zijn.B2Safe: Replicatie service om data persistent en veilig op te slaan op meerdere locaties. Data krijgt een persistent identifier. Dit is vooral handig voor communities die hun eigen repository beheren.