European Conference on Digital Libraries, 2004
This text chronicles my experiences at the European Conference on Digital Libraries , September 11-15, 2004, Bath, England.
Saturday, September 11
Taking the train from London I had a very nice discussion with a fellow passenger about art. She was a world traveler and quite a pleasant conversationalist. I arrived in Bath around noon. In an effort to aculmate myself to the time, I forced myself to stay awake and see the sights. Bath is beautiful. I toured its tiny streets, got caught in a rain shower, saw the Royal Crescent, and made my way to the University where I was staying. That evening I dined at an Italian restaurant inside a mall that was outside a public library. Alas, I never actually took the busman's holiday to the library itself.
Sunday, September 12
Since my workshops ( one on open source software and the other on XML ) were cancelled due to lack of interest, I trained back to London on Sunday. There I saw the Tower Bridge, the Tower of London, the "Gherkin" building, the British Library, and St. Paul's Cathedral. While the Tower of London is a true castle, it does not measure up to the other castles I've seen. It is a shell of its former self. The crown jewels were a bit of a disappointment too. I expected more of them. I think I enjoyed the ravens of the Tower most of all.
As I was leaving the Tower, much to my surprise, I spied the Gherkin. I had been told by my artist friend it resided near Waterloo, but she was mistaken. The building is very interesting being shaped like an elongated egg covered in blue and grey glass. I thought it was fun to look at but a good photograph was hard to capture.
I suppose I could have explored more of the British Library, but I was a bit pressed for time. While there I saw a Gutenburg bible, the Magna Carta, and a number of original Beatles manuscripts. I would have retrieved my WAIS book, but the stacks were closed.
That evening, after returning to Bath, I dined with a few new acquaintances who were also attending the conference. We discussed the European Union over Indian food.
Monday, September 13
This was the first official day of the conference, and it was filled with the usual keynote address and presentation of papers.
Tony Hey (UK e-Science Programme) described his efforts at creating networks facilitating greater collaboration. Oftentimes these networks, grid computing networks, are used for sharing data and information between scientists. As the amount of information that gets passed back and forth increases (he's talking "petabytes" or 1,024 terabytes of data!) so does the problem of curating the data and classifying it. Alas, I did not hear Hey articulate any particular solution to his problem.
Ian Witten and Katherine Don (University of Waikato) described Greenstone 3 . Their application builds on the success of the Greenstone digital library software. The entire thing is now written in Java, and it streams XML between various modules to call functions and process data. METS is now the underlying data structure, and the Greenstone team has made every effort to build backwards compatibility into the system in order to bring older Greenstone implementations along.
Giuseppe Amato (ISTI-CNR) is striving to re-use digital library applications and combining heterogeneous data/formats by mapping existing XML schema to a centralized schema. His example was MILOS (a multimedia content management system). The model consists of a number of data stores (file systems, XML streams, RAIDS, etc.) below systems that map data descriptions to centralized vocabularies. These centralized vocabularies are then used as the basis of digital library applications.
George Pyrounakis (University of Athens) outlined a digital library framework for managing his collections of folklore. Things like DSpace and Fedora were compared and contrasted. Pyrounakis plans to write a number of Web Services applications against Fedora to satisfy his need, and they are thinking about using Lucence as their indexer/search engine.
MacKenzie Smith (MIT) gave update of DSpace and encouraged the audience to think hard about the meaning of open source software. While she advocated open source software and its mechanisms, she challenged the audience to articulate its service implications. As per my own experience, she experienced very little giving back to the community, and maintaining open source software is somewhat of a drain on the hosting institution.
Dongwon Lee (Pennsylvania State University) tackled the problems of name authority control through an application called OpenDBLP . Lee recognized the problems of name searching in various databases and demonstrated the problem through applications such as CiteSeer. He classified these errors as syntactic and semantic variations, and correctly assumed that there were too many of these errors to fix through human intervention. He then classified the types of changes that names go through and boiled them down to splits, merges, and changes. Finally, he proposed putting names into a database and assigning each name an alias. It would then be this alias that all records would point to (similar to DNS names and IP addresses). Changes to the names would not change the aliases, just the underlying data. An extra advantage of this technique was the ability to travel forward as well as backwards in time to see how the names had changed.
Gils Falquet (University of Geneva) advocated building ontologies between the parts of different electronic books. By building these ontologies it will be possible to create relationships between parts and build new books and new navigation structures based on these relationships.
Bill Janssen (Palo Alto Research Center) explained some of the things he has been learning about CAR (computer altered reading) and upLib (a personal library application). The thrust of the discussion surrounded the size of page images. Users of upLib are able to upload documents to a personal storage space. Items in their collections are then browsable via icons. After much experimentation, he and his fellow co-worked decided to use a logarithmic scale to resize images accordingly. He also advocated the use of PNG files over JPEG files since PNG's compression mechanism is superior.
Ed Fox (Virginia Tech) was far and away the most busy presenter at the conference. Here he described Citiviz , a visual user interface to the CITIDEL system. Through a series of charts, graphs, radiating illustrations, and things that looked like stacked coins Citiviz attempted to illustrate the results of searches. Numbers of usability tests and timing tests where employed. Overall the additional interfaces where useful, if not just a bit confusing.
The formal discussions ended with a panel discussion surrounding the Semantic Web . I got the feeling most of the panelists thought the Semantic Web is a distinct possibility, but its success will hinge on file sharing techniques, social networking, and trust. The idea of trust then became an underlying theme for the rest of the conference.
That evening conference attendees enjoyed wine and nuts in the Roman baths. After I collected water from the baths themselves, a waiter asked me where I got the water. I confessed knowing that if he touched the bottle it would have been warm and my deception would have been revealed. He told me that he water was very unhealthy and he would have to ask his boss about what to do. Presently he returned, told me he could not allow me to keep my water, but he would be more than happy to give me some water from the originating spring. All the while I was enjoying conversation with Fabio Simeoni (Strathclyde University) and Hussein Suleman (University of Cape Town) on the topic of popular network protocols such as OAI and SRU. We dined together and I had sardines, the largest I have ever had in my experience. Drinking wine in the Roman Baths was much like drinking beer behind Roch Castle. Heady.
Tuesday, September 14
Again, this day included keynote addresses and paper presentations.
Edmund Lee (National Monuments Record Centre) had an interoperability goal regarding historic monuments. He advocated looking at digital libraries from an ecological point of view -- digital libraries require biological/human balance. There are more to digital libraries than collections and searching; there are emotional realities as well. He also described FISHXML, the vocabulary he uses to markup information about the national monuments like Salisbury Hill.
Ed Fox (Virginia Tech) described an anthropology digital library application called the ETANA-DL. He reused many parts of other digital library implementations to create something that would work specifically for anthropologists. (A primary difference between anthropologists and other scientists is the fact that anthropologist gather information in the field where there are no computers.)
Ann Apps (University of Manchester) described zetoc SOAP. This is a z39.50 current awareness application wrapped up in SOAP requests. The concept behind this application is very similar to the alerting service outlined in Ockham .
The morning's panel discussion surrounded usability and user-centered design. Things like culture, cross-cultural usability, language (specifically non-English languages), diversity, learning styles, computer accessibility were all hot topics. The one that generated the most discussion was the non-English language issue. The consensus seemed to be that websites ought to be multi-lingual in order to be most usable to the most people. Language embodies culture, and if 'sites are in English only, then subtleties of the specific cultures will come through in the implementation and non-English ideas will not necessarily get communicated effectively.
In the afternoon I listened to Kurt Maly (Old Dominion University) share his experienced with Kepler , a personalized OAI repository application. His goal is/was to share metadata describing the papers of researchers as easily as possible. He wanted to design and implement an application that was extraordinarily easy to set up, configure, and provide access to OAI repositories on people's desks. By combining an application with XML middle-ware he was able to accomplish his task. Since much of this data will reside on computers that travel around and have dynamic IP addresses, his application had to take this into account, and it did so through the use of aliases.
Mark Kornbluh , Micheal Fegan , and Dean Rehberger (Michigan State University) all shared their experiences with MATRIX, and system allowing people to locate multimedia content, enhance it with their own metadata, and re-use it in places such as other HTML pages. The goal was to expose and make use of a greater number of media files for teaching.
That evening we had wine in the art museum and listened to a string quartet play background music. On the way there I had a nice chat Ann Apps about her directory of library collections and services. I learned that their purpose and technique were very similar to the proposed purposes and techniques of Ockham. There is plenty of room for collaboration in this regard. After about an hour us conference attendees were ushered out of the museum and around the corner to an official city building where we were served an elegant meal in a large room with an almost three-story high ceiling. I sat between a librarian from Croatia who was spending time scanning catalog cards and another woman who was a technology leader. With the librarian I discussed Croatia and its geography. With the administrator I discussed the state of technology experience in most librarians.
Wednesday, September 15
Since this entire trip was being sponsored by Infomotions, I had no qualms about being a tourist yet again. At the car rental agency, when I was about to be given the keys, the clerk said, "All you have to remember is to press down when shifting into reverse." It was then I learned I was being rented a car with a manual transmission. I asked for another car and was finally given a very large Land Rover Discovery. As I drove around the country side I was the truly ugly American, dominating the road. Everybody seemed to get out of my way.
I made my way to Stonehenge. I must admit that I was impressed, but I expected it to be bigger. On the way to Avebury I got lost, and after turning around I spied a canal just waiting to be photographed. I did so, noticed a pub right there, and decided to eat. I had fish and chips beside the canal. Avebury is more impressive than Stonehege. This large circular ditch was apparently built more than 3,000 years ago for ceremonial purposes. Inside the ditch are a quite a number of standing stones and houses built during the Middle Ages. Sheep graze among the rocks. I also climbed to the top of Salisbury Hill, a very large, man-made mound of unknown origin and purpose. Some think it is a burial chamber. Others think it could be a watch tower. Nobody really knows.
Backtracking, I went to Wells. When I arrived folks were cleaning up from market day. I went to the Cathedral, and appreciated the fact that the light was coming directly from the west, and the church's interior was exceptionally illuminated. My last stop was Cheddar Gorge, a deep slice in the Earth that a person can drive through. I didn't see the dragon who is said to live there. Maybe next time. In the evening I at the "oldest house in Bath", had Welsh Rarebit, and drank lots of water.
The event was expensive but eventful.
Besides experiencing the things Bath and the surrounding areas had to offer, I learned how pervasive Web Services are in current digital library implementations. Each presented paper had the same format: motivation statement, model description, implementation, application of metrics, and summary with suggestions for future research. In almost every case, the implementation was done through the use of some sort of Web Service where one application sends another application a URL or stream of XML, the receiving application uses the URL or XML as input, processing takes place, and the results are returned to the requesting application in the form of an XML stream. Such a technique is platform independent and lends itself very well to componetized digital library services and collections. (The word "modular" seems to be out of fashion.) If such a methodology gets a foothold in the library vendor world, then the profession will need to rely on single vendors for all their applications less and less. Consequently, the "integrated library system" will be seen as a dinosaur.
I did not come away from the conference with a whole lot of new digital library knowledge. Rather I came away with a re-enforcement and re-affirmation of what I already knew. Such is not a bad thing.
Creator: Eric Lease Morgan <email@example.com>
Source: This text was never formally published.
Date created: 2004-10-01
Date updated: 2004-11-27
Subject(s): travel log; ECDL (European Conference on Digital Libraries); Bath, England;