This was the first time this event was held in the majestic and architecturally impressive city of Budapest. It was organised by The Computer and Automation Research Institute of the Hungarian Academy of Sciences (MTA SZTAKI)  and held at the Europa Congress Centre.
The event brought together a very mixed group of people from computer scientists, researchers, librarians, professors and managers. There were over 200 participants, from 36 countries. There were a total of 119 full paper submissions of which 36 were accepted after peer review, giving an acceptance rate of 30%. Also 24 poster / demo submissions and another 15 papers from the full paper submissions were accepted for poster presentation and publication in the proceedings volume. The papers were organised into the following areas: ontologies, digital libraries and the Web, models, multimedia and multilingual digital libraries, grid and peer-to-peer, preservation, user interfaces, document linking, information retrieval, personal information management, new Digital Library applications, and user studies. The proceedings from the conference are available on the Web , although access may be restricted and dependent on the subscription that your institution or organisation holds.
For those that arrived early at the conference they we were able to attend all day tutorials on the Sunday. Brief reports follow on the sessions we were able to attend.
The morning session examined the structure and use in knowledge-based assistance to users using thesauri and ontologies in digital libraries. There was a general introduction, looking at some of the main challenges for digital libraries, e.g. improving the retrieval effectiveness to handle the sheer mass of material and how thesauri, ontologies and taxonomies are needed to support such functions together with an overview of thesauri functions. There was also a look at thesaurus structure, the implementation and evaluation of thesauri (using Yahoo's classification as an example), resources available and then other examples of classifications and thesauri e.g. Medical Subject Headings (MeSH) and the Art and Architecture Thesaurus (AAT).
The afternoon session looked at the design, evaluation, and development of systems based on thesauri and ontologies. It examined the process of thesauri construction and how a conceptual structure is developed by examining facet analysis and the rules for the selection of preferred terms and concepts as descriptors. Then we carried out an activity based on developing a conceptual structure using facets. The session concluded by examining the structure and processing of thesauri data, specifically the interoperability of thesauri and ontologies, the structure of a thesaurus/ontology database, the forms of Knowledge Organisation Systems and a look at thesauri software and evaluation.
The organisers addressed various key issues facing large-scale digital library infrastructure that have been examined in two EU-funded initiatives (the BRICKS Project  and the Digital Library Management System strand of the DELOS Project . One of the goals of their work which they presented has been to investigate the possibility of dynamic digital library infrastructure. This has fallen into four main areas: the development of a Service-Oriented Architecture approach to the creation of digital library services, peer-to-peer hosting and discovery of such services, the development of a tool to support the dynamic creation of virtual collections, and the development of the high-level software tools needed to support the three other goals. The tutorial explored the issues and reasons for these goals and local versions of the prototype tools were demonstrated.
The second half of the tutorial addressed two main areas: the application of the 5S model  and work on the development of a digital library curriculum . A case study was presented that used the 5S model to support interoperability mapping between a group of archaeological digital libraries. Other applications of the model presented included using the model to examine repository quality and create a template to install and configure DSpace. The work on the development of a digital library curriculum followed on from this neatly and presented the progress of a NSF-funded project to map out a course of study on digital libraries. The composition of the curriculum has been based on numbers of papers presented at key digital library conferences and articles in targeted journals. It should become a key point of reference for educators interested in the area.
There were two keynote sessions this year, delivered on the mornings of Monday 17 and Tuesday 18 September.
Seamus Ross, Professor of Humanities Informatics and Digital Curation, and Director of Humanities Computing and Information Management at the University of Glasgow, runs HATII (Humanities Advanced Technology and Information Institute) of which he is the founding director. He is also Associate Director of the Digital Curation Centre in the UK (since 2004), a co-principal investigator in the DELOS Digital Libraries Network of Excellence (since 2002), Principal Director of Digital Preservation Europe (DPE) (since 2006).
Seamus gave an interesting overview of the use of preservation (starting from the 17th century) to the role of digital libraries (especially over the past twenty years) in digital preservation. He argued that digital libraries, whether commercial, public or personal, lie at the heart of the information society. He then went on to demonstrate that there has been little research into the long-term sustainability and meaningful accessibility of content from digital libraries. He also presented the case that much more work was needed on the actual theories, methods, and technologies that can either foster or ensure digital longevity. Seamus outlined the key research challenges, theoretical, methodological, and technological requiring attention from researchers in digital libraries during the coming years. Seamus concluded that digital preservation in digital libraries needs to be based on a theoretical framework of archival science if DLs are to retain long-term viability at the centre of the information society. The full transcript of his presentation is available .
Professor Sølvberg is also Dean of NTNU's Faculty of Information Technology, Mathematics and Electrical Engineering. He is chair of the Board of Wi-Fi Trondheim. On the second day of the conference Arne gave a keynote presentation on Wi-Fi Trondheim , which allows users full Wi-Fi access throughout Trondheim. Arne Solverberg described how they had 'carpet-bombed' Trondheim with wireless devices so that as many people as possible could have wireless access. This was a very impressive demonstration of collaboration funded between the University, the City Council, the County, the local bank, the local electricity provider, and the local newspaper. He envisaged that broadband connectivity to everyone in Trondheim (population over 161, 000) would be a reality by 2009. Arne then went on to present some of the technical, business, regulatory, as well as service provision challenges involved. One interesting question focused on the reaction of local Internet Service Providers to this project, and how threatened they felt by it.
In addition to the two keynote sessions, there were several parallel sessions that were delivered over 3 days that covered the following topics: ontologies, digital libraries and the Web, panel discussion on the experiences of several European projects examining different aspects of digital libraries (BRICKS, TEL, MICHAEL and DELOS), as well as topics such as Multimedia and Multilingual Digital Libraries, Grid and Peer-to-Peer, Preservation, User Interfaces, and Digital Libraries in Central and Eastern Europe. Mention should also be made of: Infrastructure Challenges for the New Europe, Document Linking, Information Retrieval, Personal Information Management, New DL Applications and User Studies. Of course we were not able to attend all the sessions and you are able to see the full papers in the proceedings of the conference . Instead, we have highlighted a few presentations that caught our eye over these three days.
One particular presentation by Riccardo Miotto and Nicola Orio on the Automatic Identification of Music Works through Audio Matching was fascinating. The project looked at thousands of recordings of rehearsals for classical concerts that have no metadata. The study was trying to identify 'fingerprints' of performances (for which there are metadata) and trying to use them to match metadata with recordings of rehearsals where no descriptions exist, therefore automatically classifying them. The presenters also discussed possible commercial applications of their work.
One paper in particular stood out, partly because of the catchy title: Opening Schrödingers Library: Semi-automatic QA Reduces Uncertainty in Object Transformation. This paper, by Lars R. Clausen of the State and University Library in Århus, Denmark, describes an approach to semi-automated quality assurance and applications designed to support the goals of preservation.
The poster session began by employing the successful format in which each poster presenter is permitted a few sentences describing their work - referred to in some conferences as 'one-minute madness'. Posters showcased ideas, developments and work in progress from various fields, from social tagging to repositories and services. However the primary value of the session for participants was probably the lively discussions that took place between participants, and with delegates in general. 38 posters and demos were accepted for the conference, a full list is available .
This session contained three papers on the theme of novel applications in the digital library sphere. The first described an OCR-free approach of establishing layout similarity measurements for pages, information which can be employed as a novel retrieval metric. The second discussed assessment techniques for data integrity in the scientific digital library context. The third, by Hussein Suleman, discussed the Bleek and Lloyd collection, a set of books and drawings that document the language and culture of some Bushman groups in Southern Africa, which was designed to employ no database technologies due to preservation concerns - XML-based records were employed instead. Scalability was described as a principal limiting factor in employing this strategy.
One particular presentation stood out for its applicability to everyone at the conference and anyone that uses a computer with a desktop. This was the presentation on the way we organise our desktops and folder structures, 'Personal Environment Management' by Anna Zacchi and Frank Shipman from Texas A&M University. It was clear that this was of interest to the whole audience and the results showed that there is great variance in the way we organise our desktops. The presenter showed very different desktops, on a range of different operating systems (Windows, OSX and Linux). It was fascinating to see how much one's desktop tells us about the individual; one would not necessarily wish to have one's own desktop so analysed!
The official conference dinner was held on a cruise ship on the blue Danube, although at that time of day the river in fact looked very nearly black.
As is tradition at ECDL, the official conference finale involved presentations and awards to the best paper and best poster.
This year the winner of the best paper was Investigating Document Triage on Paper and Electronic Media by George Buchanan and Fernando Loizides from the Future Interaction Technology Laboratory, University of Wales, Swansea. Document triage is the decision that is made in information-seeking when the user first decides on the relevance of a document to the information required.
The award of the best poster was shared jointly between:
Location and Format Independent Distributed Annotations for Collaborative Research by Fabio Corubolo, Paul B. Watry and John Harrison of the University of Liverpool and EOD - European Network of Libraries for eBooks on Demand by Zoltán Mezõ, Sonja Svolšjak and Silvia Gstrein of the National Széchényi Library, Budapest, Hungary, National University Library, Ljubljana, Slovenia and the University of Innsbruck, Austria, respectively.
The repository ecology workshop was organised by Robert John Robertson and Mahendra Mahey. The purpose of the workshop was to enable delegates to explore how the ecology metaphor can be applied to examine how users, repositories, and services interact within an information environment. The concept of a repository ecology was presented along with illustrative examples of repository ecologies drawn from different domains. Throughout the day there were also presentations from researchers about their own understanding and interpretation of the repository ecology from their domain or area of work. There were presentations from:
Delegates in the workshop were given very practical activities which involved the identification of: components in a repository ecology; opportunities for interaction; and gaps in the ecology. Delegates were then encouraged to produce a model of their ecology and present it to the group, as can be seen below:
Further information about this workshop is available .
This year's Networked Knowledge Organisation Systems (NKOS) workshop was the sixth European event in the series, which has been held at ECDL in the years 2000, 2003, 2004, 2005 and 2006. NKOS events have also been held concurrently with JCDL and DCMI conferences. The programme covered a number of topics. The day began with several presentations in the area of social tagging, including a literature review in the area by Ali Shiri and two pieces of work in the area of exploration of patterns of tag use.
Next came a panel on the topic of ISO NP 25964 - structured vocabularies for information retrieval, revising existing international standards for monolingual and multilingual thesauri. The third section of the day was entitled Implications of online KOS applications, and brought together a discussion of facet analysis by Vanda Broughton and a case study by Michael Panzer discussing the 'webification' of controlled vocabularies - so that vocabularies are exposed online as building blocks intended for use and reuse.
The afternoon was principally employed in discussion of semantic mappings between vocabularies and terminologies, together with a presentation by Mateusz Kaczmarek and colleagues, who discussed the possibility of collaboratively building crosswalks between terminologies.
The conference was very well organised, and special thanks must go to Magdolna Zsivnovszki and her team from the Department of Distributed System, Computer and Automation Research Institute, Hungarian Academy of Sciences. Conferences like this offer a great opportunity to network with people from varied backgrounds and countries. Our only regret was that we did not get a chance to see the very impressive- looking city of Budapest. Next year's conference will be held in Aarhus, Denmark from 14 - 19 September 2008 and will be organised by the State and University Library and Åarhus University .