Search   |   Back Issues   |   Author Index   |   Title Index   |   Contents

Conference Report


D-Lib Magazine
April 2004

Volume 10 Number 4

ISSN 1082-9873

Report on the 3rd OAI Workshop

CERN, Geneva, 12-14 February 2004


Chris Awre
Development Group
Joint Information Systems Committee (JISC)

Red Line


This 3rd OAI Workshop, entitled 'Implementing the benefits of OAI', was held at CERN on 12-14 February 2004. CERN has also been the location for the first two workshops and this third workshop took place 18 months after the last one. The growing interest in OAI and the open access movement led to a big increase in the number of delegates to the workshop (from 120-190) this year and also the number of countries represented (27 this time). The workshop itself comprised a number of presentations, four tutorials, seven breakout groups and a panel discussion, plus plenty of time for discussion and interaction between fellow delegates.

The full agenda, complete with Powerpoint slides and video of the talks where available, can be found at <>. This report highlights some of the main points raised and the conclusions reached by the workshop. For full details, please refer to the agenda.

Two keynote presentations were given. The first, by Diann Rusch-Feja, Director of Information Resources at the International University in Bremen, considered the current state-of-play in the relationship between OAI and scientific publishing. Her talk presented many questions that currently need to be asked about this relationship and provide a good checklist of issues that are outstanding. OAI is encouraging the dis-integration and dis-location of scientific publishing and uptake of the standard is increasing in different sectors, e.g., the library portal market. The openness brought about by use of standards such as OAI is also spreading, with initiatives such as the Connexions project at Rice University (see in the US creating a 'content commons'. This was used as an example of a cross-institutional composite server.

Carl Lagoze gave an update on development of the OAI itself. Current projects involve an OAI registry allowing repositories to be discovered, ERRoLs, an OCLC project on OAI identifiers, OAI Static Repositories, and OAI-rights, developing the outputs of the RoMEO project ( into a generic solution for the OAI. Some non-traditional uses for OAI are also being examined: using OAI for the disclosure of web usage logs; providing multiple service access to locally hosted assets using OAI—developing the idea of a repository architecture; OAI and P2P; and the use of RDF within OAI. It was clear that there is no shortage of ideas!

Chris Awre from JISC gave an overview of the FAIR programme and uses of OAI within the UK. There is much use within the e-print and e-theses communities, but also a great deal of interest in museums and image collections for disclosing their content, and also interest in how OAI might support preservation activities. This was followed by a similar overview of the DARE programme from Holland by Lilian Van der Vaart from SURF. The DARE programme has put in place institutional repositories at all Dutch universities and has used sample content to test out the infrastructure; a demonstrator showing the different repositories, which use a variety of software systems, can be seen at <>. DARE is now moving to investigate how these repositories can be filled through regular deposit by academics—one idea for promoting this will be to tender for small projects to show what can be done with documents and metadata placed in a repository.

Peter Schirmbacher described the work of DINI, the German Institute for Networked Information (, though no English version). This membership organisation was only formed in 2002 and is still establishing its role. However, it has carried out a survey of repository usage in Germany and has put together a set of criteria for a repository. Adherence to these criteria permits the award of a certificate from DINI, which intends this as a quality standard. There is a lot of interest in using institutional repositories for e-publishing within Germany, which appear to parallel similar developments in Australia.

Philip Hunter gave an overview of the OA Forum and the OAI tutorial that had come out of this work (see He also described the e-Bank project, connecting e-prints to research data. Theresa Veldren described the initiative within the Max Planck Society to make research outputs available on open access. This work has led to the creation of MPG eDoc, a repository software package that came out of the gathering of publications for their annual report. The Society view the journals crisis as a symptom of an outdated system of dissemination and scholarly communication, especially in the era of the Internet and emerging eScience. They have also recognised that open access is a long-term commitment and are prepared for the long haul. Theresa reported that there is a follow-up meeting to the Berlin Declaration being planned in 2004, to investigate how the principles laid down by this can be implemented.

Martin Wynne from the Oxford Text Archive described the work of the Open Language Archives Community (OLAC) and their use of OAI ( Their developments have included work with multiple content types, sharing of metadata with different service providers, extending Dublin Core to include specific elements for languages, making use of the OAI static repository specification, and embedding access to the Archive in other sites. Colin Steele from the Australian National University then showcased repository developments at Australian universities, including the interest in e-presses and the recent development at Queensland University of Technology of mandating that all research outputs should be placed in the institutional repository. Lastly for the first day (all 10 talks of it!), Ziga Turk described the EU SciX project ( that has developed a system to facilitate electronic publishing of any content.

Day 2 kicked off with a view of the pros and cons of open access. This highlighted that although there are key reasons why open access is better, there are large, predominantly cultural, barriers that need to be overcome to allow this to happen. The SPARC Next Steps program, to be developed this year, was briefly described. Christiane Asschenfeldt, the International Co-ordinator for Creative Commons, described the work of Creative Commons and its internationalisation, details of which can be found at <>.

David Prosser very neatly demonstrated how the two strands of open access, self-archiving and open access journals, can and indeed need to work together. Both need development in parallel to fulfil the role that journals currently have. Reference was also made to the recent OECD Declaration on Access to Research Data from Public Funding (,2340,en_2649_201185_25998799_1_1_1_1,00.html—scroll down to past the minutes of the meeting). Lotte Jorgensen described the latest developments at the Directory of Open Access Journals (DOAJ— at the University of Lund. Now established, the Directory is aiming for greater reliability, content, impact, and communication. Article level access is currently being planned. Interestingly, it was noted that not many open access journals were harvestable using OAI as yet. David Yakimischak and Michael Krot from JSTOR then briefly described the changes being introduced to the JSTOR service in 2004, which will include support for OAI-PMH. The UML diagrams used to construct the architecture for the new service will be made available for wider use as well.

Bill Hubbard described the UK's SHERPA project with particular reference to the strategies to be used to encourage academics to deposit within institutional repositories (which the project is calling institutionally-based e-print repositories or IBERs). As part of this he referred to the P problems: publicity, politics, promotion, preservation, presentation. Saskia Franken described the Igitur publishing and archiving services at the University of Utrecht ( As part of the DARE programme, they have adopted a number of steps to getting content: get buy-in from senior management; publicity; just start; collaborate with others; make 'publishing' (akin to self-archiving) easy; make the benefits clear; and offer special added value services around the addition of content to the repository. Rudiger Voss from CERN then described the two-stage peer review system employed by particle physicists (particularly relevant at CERN, of course). The final talk of the day came from Thomas Krichel by telephone, as visa restrictions had prevented his attendance. He proposed that institutional repositories should be backed up by discipline-specific aggregators, as most academics were more attached to their subject than their institution. Examples already in existence included NCSTRL, CORR, CiteSeer and DBLP. His own konz project ( is based on the last of these. This seeks to search the Internet for full text based on bibliographic titles, potentially removing the need for a formal repository! His desire is to see scholarly communication more author-driven. As part of this, he proposed the issuing of certification for putting articles in repositories and appealing to an author's vanity.

Seven breakout groups took place during the workshop—see the agenda for details. These were reported on the Saturday morning and the key points from each were as follows:

  1. Implementation: the FAIR and DARE experience
    The role of the library as publisher and whether this would represent competition or complementarity with publishers was raised as a major issue for libraries to consider when implementing an institutional repository. It was also stressed that there needed to be full integration of institutional repositories with learning, teaching and research.
  1. The relationship between OAI-PMH and Dublin Core: Required, recommended or other?
    In discussing the use of Dublin Core with OAI, there was much discussion around the quality of the metadata being entered, particularly by non-qualified people (i.e., most self-archivers). There is also limited time for such metadata creation. With regard to using other metadata formats, it was suggested that the mandatory link to DC be kept, but that DC.type and DC.identifier should be used and then the record linked to the richer metadata format. It was imperative that any solution should not invalidate previous work.
  1. Possible Enhancements to OAI-PMH including MetaSearching and Authentication
    Discussion of metasearch highlighted the possibility of using OAI-MH particularly if it could be extended. Authentication to open archives might also be useful to prevent abuse and to be able to offer individual responses.
  1. Overlay Journals
    Overlay journals, constructing a 'journal' out of articles in one or multiple repositories, offer a way to certify or quality mark items—they can be used as a selling point. Cross-institutional overlays were tricky without widespread institutional repository infrastructure and clear metadata support. There is also an identifier problem.
  1. The OAI-PMH community: How to enable better collaboration and communication between Data and Service Providers
    It was concluded that OAI data providers and OAI service providers aren't talking to each other enough in a discussion on these two. As such there is a lack of standards, which makes effective service provision difficult. Registries of metadata crosswalks would facilitate interaction.
  1. Open Access Citation Index
    A new working group to examine citation indexes for open access materials was set up after discussion of this matter.
  1. If You Build it, Will They Come? Filling an Institutional Repository
    Possible routes to filling institutional repositories included: working with liaison librarians more; education on copyright; pushing preservation; and encouraging academics to discuss alternative publishing models with their Learned Society.

The workshop finished with an expert panel that gave a range of views on open access, from totally in favour through to strong scepticism. The panel highlighted the many different views in the communities we are working with, and the long path ahead if any change is to come about. Some of the main views expressed by the members were as follows:

Peter Suber, Open Access advocate - Stressed the need for journals to stop preventing publication on the grounds of prior publication (in a repository); there is a need for deposit to be made a requirement of funding; and that open access journals should deposit their content in OAI-compliant repositories.

David Prosser, SPARC Europe - Also stressed that funding agencies will be the main driver in encouraging open access.

Bas Savenije, Librarian - Implied that the term 'publisher' may become defunct as the roles become spread over many parties.

Simeon Warner, OAI developer - Content, not quality is the key to institutional repositories. There is also a need to get open and paid-for resources on a more level playing field.

Ian Butterworth, physicist and member of UK research council committees - Is not keen on the closeness of institutional repositories and open access journals and prefers subject repositories for research—institutional repositories are fine for other materials. Greater proactive advocacy is required for whatever route is chosen.

Howard Flack, editor of Learned Society journal - Regards open access in terms of fair use. The Learned Society publishing model has worked extremely well for them and they see no reason to change. They would also be concerned about abuse of OAI and someone re-constituting the journal.

Desmond Reaney, Institue of Physics Publishing - Professed uncertainty that open access is viable in the long-term. Open access doesn't suit the author's needs. Authors want authority—which route will offer them this best?

Overall, the workshop concluded that technically the tools are in place for extensive dissemination of materials on an open access basis. Implementations are now common. However, the wider issue of bringing contributors on board is not yet won and much work is still required before the technical solution can be considered an overall solution.

A 4th OAI Workshop is planned for October 2005.


Copyright © 2004 Chris Awre

Top | Contents
Search | Author Index | Title Index | Back Issues
Previous Article | In Brief
Home | E-mail the Editor


D-Lib Magazine Access Terms and Conditions

DOI: 10.1045/april2004-awre