2. Information Technology Services

3. CRIS Systems and Open Access Repositories 3 Dr Mark Cox Research and Learning Development Manager, King’s College London mark.cox@kcl.ac.uk

4. CRISes and OARs CRISes and OARs developed substantially separately OARs primarily developed as discovery tools – Public facing – Generally populated via individual deposit – Development usually led by library staff CRISes mainly used as (data) management tools – Used within institutions, with little or no public interface – Often built to integrate with other systems to allow import of data – Most implementations led by institution’s research offices Considerable overlap in the data that is collected by each 4

5. Integration of CRISes and OARs Increasing realisation in institutions that there is benefit from streamlining OAR and CRIS development Several potential ways of doing this Can use a CRIS for gathering all research related information, including research output, and feed/link to repository Can expand repository to be able to cover a wider range of research related information Can use portal functionality of sophisticated CRIS systems to mimic repository functionality 5

6. Research Information systems at King ’ s 6 Web interfaces Research Gateway Database HRMS Name, job title, start date, etc Feed automatic, overnight SITS PG(R) details Feed automatic, overnight APTOS Grant monthly expenditure Feed automatic, overnight Res. Grants & Contracts Awards, amounts, dates Feed manual, 2 weeks IoP Res. Grants Awards, amounts, dates Feed manual, 1 month Experts database (potential) Staff Profiles (Web pages) RAE/REF outputs AKORD publications database (web interface) InCites -bibliometric information RG Academic Interface Manual entry: Qualifications Esteem Factors Publications not in WoK feed Feedback workflow RG Admin Interface Manual entry: Qualifications Esteem Factors Publications not in WoK feed Feedback workflow AKORD Repository Full text of publications, publication metadata Web of Knowledge Publications in journals Feed automatic, 1 week School web pages (potential) School/College Managers – Research Strategy Direct link

7. euroCRIS euroCRIS nurtures the CRIS community by events, the euroCRIS website, a newsletter, online discussion forums and other appropriate mechanisms. Members, strategic partners and external experts meet in a series of biannual membership meetings, annual strategic seminars and biennial conferences. The output from these events is publicly available. euroCRIS are the curators of the CERIF standard http://www.eurocris.org 7

8. CERIF Common European Research Information Format International-standard data model which describes Research Entities and their relationships Developed originally with the support of the European Commission Development now under the custody of euroCRIS CERIF is neutral to architecture 8

9. CERIF data model 9

10. Benefits of using CERIF 10 Institutions do not have the resources to develop a completely new UK data model CERIF has been successfully used as the basis for a number of national systems across Europe CERIF has evolved over ten years via the experience of developers across Europe euroCRIS, as the custodians of the CERIF standard, form an active and enthusiastic community which encourages participation in the development of the model

11. REF Three major factors – Research Outputs – Environment – Impact (new to REF) Use of bibliometric data still under discussion Major driver for development of systems 11

12. RAE 2008 submission Some use of repositories to gather information on research outputs Some use of Research Information Systems (CRISes) Large amounts of manual entry Most CRISes at the time were bespoke developments CRISes usually held information on publications Little consistency across different institutions (and sometimes within institutions!) 12

13. Towards the REF Most institutions now have an Institutional Repository, usually managed by library staff Some institutions have in-house built CRISes, sometimes integrated with other systems Several institutions have purchased commercial CRIS systems As the requirement for data relating to research increases, many institutions are looking for a more co-ordinated approach to research information management A growing appreciation that collaboration and harmonization will provide greater value for money, both for institutions and for UK research in general 13

14. Readiness for REF (R4R) 14 The R4R Project between King’s and Southampton is examining CERIF as a the basis of a mechanism for delivering data to the REF At the core of this will be a “lightweight “ data schema based on the CERIF model, adapted to the needs of data collection for the REF – “CERIF4REF” The desired outcome is that the CERIF data model will: – form the “glue” that will hold disparate information systems together – allow for interoperability and exchange of data within and across institutions, including external systems – deliver repository and CRIS interoperability for the REF

15. R4R work streams 15 Carry out an analysis of current practice in UK institutions From REF requirements, develop the CERIF4REF data schema Evaluate CERIF4REF schema applicability Examine mapping this data model to external sources Create “plug-ins” for the three most common types of repository (ePrints, DSpace, Fedora) Demonstrate data exchange between systems at King’s and Southampton

16. 16 CERIF4REF Involved an iterative mapping exercise, reviewed and refined by euroCRIS developers Vast majority of RAE data fields could be mapped to CERIF elements Missing DoB element was incorporated into latest version of CERIF Certain RAE data fields (especially in RA3 and RA4 forms) required aggregation of lower-level CERIF elements

17. 17 CERIF4REF schema To aid with defining elements, an XML schema (.xsd) has been created for CERIF4REF, accompanied by a data dictionary This describes the underlying elements that are required to generate RAE/REF summary data, such as RA3 and RA4 forms Stylesheets (.xsl) have also been generated which can be used to transform CERIF4REF data into pure CERIF format, or into RAE forms.

18. CERIF4REF in practice 18 Full CERIF Oracle database created by using SQL scripts developed by euroCRIS CERIF4REF model also used to create separate Oracle database Data mapping to existing King’s CRIS (Research Gateway) carried out CERIF4REF database successfully populated from King’s Research Gateway system

19. Evaluating the CERIF4REF schema CERIF4REF model being used to inform development of local ePrints repository at Southampton Case studies carried out with a number of other institutions to determine if CERIF4REF elements could be populated from existing systems Institutions chosen by size and existence (or absence) of a repository Agreement that CERIF4REF covered a majority of areas 19

20. External Sources 20 Publication data (Thomson Reuters) – Tool developed at Southampton to allow population of repositories from Web of Science (WoS) basic API via the SWORD protocol – This may circumvent the need to map this data to CERIF UK Research Councils – RCUK are developing a new system to collect research output from institutions – Discussions have been held with members of this project team to determine the overlap of the requirements with the R4R developments

21. Repository plug-ins 21 Development of plug-ins for ePrints (Southampton), DSpace (Edinburgh) and Fedora (King’s) Tools generate xml in CERIF4REF format to allow import into other repositories or CRISes Focus is on research outputs Southampton are, in parallel, developing an expanded ePrints system Plugins available from R4R website at http://r4r.cerch.kcl.ac.uk http://r4r.cerch.kcl.ac.uk

22. R4R: Conclusions so far 22 In the UK, there is much greater awareness of, and interest in, the CERIF model and the role of repositories and CRISes in research management CERIF being increasingly viewed as the best solution for data harmonization within the UK CERIF4REF model has potential to provide a step towards greater CERIF uptake in UK institutions

23. MICE project Measuring Impact under CERIF Impact currently not covered by CERIF Investigation of the possible areas of impact Defining the “elements” of impact Creation of a CERIF schema to model impact Implementing impact measurement into a CRIS Website at http://mice.cerch.kcl.ac.ukhttp://mice.cerch.kcl.ac.uk 23

24. Determining indicators of impact Evaluation of a number of existing documents regarding impact (including REF impact pilots) Aim was to create a comprehensive set of impact indicators Attempt to categorise these (economic, health related, cultural, etc) Tree diagram created to show these relationships Also considered were any tangible measures of each type of impact 24

25. Proposed CERIF description of impact 25

26. Impact measurement in a CRIS Feasibility study carried out to determine how to model impact in the PURE commercial CRIS Using tree diagram from earlier work, indicators of impact can be categorised Some impact measures already exist in PURE data model, but many would need to be extensions With such a wide variety of possible indicators and measures of impact, standardisation and agreement would be required 26

27. Related Developments JISC research information management (RIM) group – following up recommendations of EXRI report that CERIF should form a common research information exchange format for the UK REF submission system likely to include some form of CERIF upload capability 27

28. euroCRIS Rome Declaration on CRIS and OAR We, representing the communities concerned with CRIS and OAR hereby declare: That high quality research information is critically important to research institutions, research funders, policymakers and society at large; That information on publicly funded research should be available, shareable and integrated seamlessly Therefore we commit: To develop, support and promote an architecture (including data model and services) suitable to pursue these principles; Jointly to adopt, develop and promote applicable open standards; To advocate these principles among all stakeholders. 28

29. Questions? ? 29