University of Bristol
Bristol BS8 1HH
ECDL-2003 Conference Notes
ECDL2003 was the seventh in the annual series of European Digital Library conferences, this year hosted in Trondheim, Norway. The unusual move from September to August does not carry through to next year's conference at the University of Bath, UK, which returns to the 'normal' September slot (12-16 September).
My interests in digital library applications, user perspectives and service management obviously influence my 'take' on the conference experience and the sessions I attend. ECDL2000 in Lisbon seemed to me to have an emphasis on technologies and hard computer science which I found difficult to engage with. The Rome experience at ECDL2002 was more 'user-friendly', with a small but interesting number of papers on user interactions with digital libraries and on applications. It was interesting to note further growth of these topics in papers presented at ECDL2003, presaged by the Rector of the Norwegian University of Science and Technology, in his opening address, noting the growing importance of interdisciplinary work for digital libraries research and practice.
A fairly high-level summary of some of the highlights of the conference is given below, with presenter/paper references where appropriate. A small bone of contention for many who present (and for those unable to attend) is the print-only publication of papers. The proceedings are available in the Springer 'Lecture Notes in Computer Science' series (volume LNCS 2769) . This year, as a move towards wider publication, some of the presentations (in PowerPoint or PDF) are also available from the conference Web site .
Conference in a Nutshell
There were two main conference tracks which reflected the computer science and more user-oriented topics mentioned above. Whilst still treated as separate, the balance between them is a step in the interdisciplinary direction. My general impression is that there was a lot of interest in RDF/XML, ontologies and Semantic Web-related languages and technologies, and not just from the computer scientists. The 'big research things' emerging from the sessions I attended seemed to be:
- More intelligent search technologies for applications and for end users; more relevant results from both structured and unstructured data
- Further development of annotation, recommendation and personalisation systems to enhance use of digital library collections
- Gaps in provision of robust digital preservation and Web archiving mechanisms
- Continuing and iterative analysis of user behaviour (in different cultural, subject and other contexts) in digital library use
There also appeared to be a willingness to question why digital library features and services were being offered, rather than an acceptance that 'just because we can, we should'. Evaluation of added value, as well as of technologies per se, was a recurrent theme.
Formats and Findings
This year, there was an attempt to introduce sessions in different formats alongside the traditional papers. The poster session , for example, was preceded by a 'one-minute madness', with 50+ poster and demo presenters giving a 60 second sales pitch with slides to encourage delegates to visit them during the drinks reception. This was both nerve-wracking (for those of us presenting) and entertaining and energising (for those of us in the audience). It could all have gone horribly wrong, but it was actually a very good way of getting a feel for the projects and ideas being presented and provided an opportunity to put faces to names.
John Lervik (FAST) in his opening keynote , listed the potential of search technologies for digital libraries of the future. His focus was on the systems (query and data analysis engines, relevance research and entity extraction, adaptive and customisable services) and the notion that 'search engines can do more than just search'. During questions, he also highlighted the importance of user interface design and user studies in ensuring that user expectations (built on the simplicity of Google and single keyword search preferences) can be met by digital libraries with such advanced search technologies.
User studies featured strongly in the subsequent paper session  which showcased experiences of user studies from three countries (France, Slovakia and the UK). A study of use of, amongst others, the national library of France's digital library (Gallica) undertaken by FranceTelecom, used a methodology that involved the installation of 'tracking' software on user computers for six months. The study also involved a number of more traditional questionnaires and generated a typology of digital library users (for Bibliothèque nationale de France (BnF)), revealed a general preference for searching over browsing and a disinclination to read online. Most interestingly, the study revealed a new audience for 'academic' digital libraries including the "non-professional researcher" interested in family or regional history.
An ad hoc panel convened at the end of this first paper session included a reminder (from Slovakia) that network access is not such a given in other countries as it is increasingly becoming here in the UK.
The 'official' panel sessions were also interesting. A clash of commercial and open access publishers on 'The future of academic publishing' featured panellists Leo Waaijers (Wageningen University Library); David C Prosser (SPARC Europe) and Michael Mabe (Elsevier Science). Given that so much has already been said on both methods with little sign of consensus, there was little new from either panellists here or from the floor.
The question: Digital preservation - are metadata really crucial? generated some thought-provoking points from the audience and panellists Michael Day (UKOLN, University of Bath); Steve Knight (National Library of New Zealand) and Catherine Owen (Performing Arts Data Service). There is already a huge amount of metadata 'out there' and we are still unsure 'how much metadata is enough'. Migration tools are not yet fully developed. Our priority now should be preservation of digital objects for the next 5-10 years rather than trying to second-guess the future.
Papers on annotation and recommender systems  focused on a different kind of discovery and interestingly noted the potential for 'undiscovery'. This can occur when suggestions are made to users on the basis of their previous patterns of digital library use (and those of others with similar expressed or inferred interests). This session also raised issues around users' willingness (or, it seemed, otherwise) to read and make annotations online.
ECDL2003 was for me a surprising success: this may be a reflection of the maturity of the field. That is not to say that there were no innovations. This is obviously still an area where there is much to be developed and researched and where there are some exciting initiatives underway. But I certainly had the impression that much of what was being discussed was in the mainstream of problems faced by digital library providers and users and offered learning opportunities for many.
The wireless network worked extremely well and did away with the usual unseemly queues for machines where people can check their email. This could also have provided an excellent opportunity for greater audience participation and exchange of views via IRC or Weblogs as I have been told happened at the last WWW conference: as far as I am aware, there were no organised channels for this and I didn't come across any informal ones either ... unless someone can tell me different?
The ECDL2004 Conference  will take place in Bath, UK, 12-16 September 2004.
- Research and Advanced Technology for Digital Libraries, Traugott Koch and Ingeborg Torvik Solvberg (eds): Lecture Notes in Computer Science, Springer LCNS 2769 Order details available from:
- ECDL 2003 Conference Web site
- List of accepted demos and posters
- Digital Libraries: What Should We Expect from Search Engines (Presentation slides in PDF). John M Lervik, CEO and Co-Founder, Fast Search and Transfer (FAST).
- Session 2A:
Uses, users and user interaction. Presentations: Users and Uses of Online Digital Libraries in France. (PDF) Houssem Assadi, Thomas Beauvisage, Catherine Lupovici and Thierry Cloarec.
In search for patterns of user interaction for digital libraries. (PowerPoint) Jela Steinerová.
Evaluating the Changes in Knowledge and Attitudes of Digital Library Users. (PowerPoint) Gemma Madle, Patty Kostkova, Jane Mani-Saada and Julius R Weinberg.
- Session 3A:
Annotation and recommendation. Presentations: Annotations: Enriching a Digital Library. (PDF) Maristella Agosti and Nicola Ferro.
Identifying Useful Passages in Documents based on Annotation Patterns. Frank Shipman, Morgan Price, Catherine C. Marshall and Gene Golovchinsky.
Others Also Use: A Robust Recommender System for Scientific Libraries. (PDF) Andreas Geyer-Schulz, Andreas Neumann and Anke Thede.
- The ECDL2004 Conference Web site http://www.ecdl2004.org
Lesly Huxley is Research Director at the Institute for Learning and Research Technology (ILRT) at the University of Bristol. In her eight years in the Institute she has worked on a number of national and European digital library projects including DESIRE, Renardus, SOSIG and Regard. She was Publicity Chair for ECDL2003 and is Workshop Chair for ECDL2004.
Article Title: "ECDL2003: Conference Notes"
Author: Lesly Huxley
Publication Date: 30-October-2003
Publication: Ariadne Issue 37
Originating URL: http://www.ariadne.ac.uk/issue37/ecdl2003-rpt/