workshop on Interoperability of Web-Based Educational Systems
The workshop on Interoperability of Web-Based Educational Systems at the 14th International World Wide Web Conference (WWW2005) in Tokyo took place today.
Dan Rehak gave an update on the current status of CORDRA. One of the important misconceptions, it seems to me, that Dan was very explicit about is that CORDRA is not a new standards initiative: rather, the folks behind CORDRA believe that the standards required are all available. Dan’s slides include some good remarks on requirements for a content repository and infrastructure.
Bernd Simon presented the work on the Simple Query Interface (SQI), a basic protocol for query transport that is rapidly maturing as an official CEN/ISSS standard. Bernd demonstrated the federated search through SQI within the ELENA project.
Stefan Sobernig talked about query translation between RDF and XML, that enables exposing XML metadata as RDF repositories, mapping datalog-QEL queries into XQuery. Both this and the previous presentation also relied on SQI.
Roberto Tedesco presented the virtual campus work on reuse. Central to their approach is the notion of recursive composition of learning objects. Roberto made some remarks about LOM: the exact meaning is sometimes difficult (semantic density), some aspects are missing (group or individual learning), and LOM records are not fully automatically processable (installation remarks is free text). They added data elements on supervision mode (none, tutored, …), cooperation, communication, whether or not learners are expected to produce an artifact, etc. There are plans to incorporate CC/PP to deal in a more flexible way with technical requirements.
Yannis Kalfoglou focused on exploiting multiple ontologies, with the specific aim of involving the community of end users, in the AKT project. When I asked how we can do this in a non-obtrusive way, then answer was that this is “still a matter of research”. Well…
Bich-Lien Doan discussed how to do complex queries that refer to the structure of content, as in “objects that are a part of data mining courses and articial intelligence courses”. Their approach relies on LOM for description of the learning objects, and an ontology that defines is-part-of relationships over which some LOM characteristics are inherited.
Liddy Neville focused on accessibility issues. (BTW, Liddy mentioned that a microsoft study apparently shows that about 60% of users have accessibility issues!)
According to the authors, WAI doesn’t address all issues, and tools are not supporting the guidelines, and therefor they wanted to start anew. User requirements are grouped in control, display and content. The accessibility properties must be described in trust-worthy metadata, and include information about the primary object, alternative objects, and supplementary objects. After the presentation, Dan Rehak, Wayne Hodgins and myself argued that, rather than separating out accesssibility issues, these should be treated through a “universal design” approach, where modularization and individualization also address for instance issues around multi-linguality and -culturality.
In the short paper section, Mitsuru Ikeda presented work on superimposing ontological information over SCORM content. The following paper dealt with the use of cultural heritage material for educational material. That was followed by a presentation on a remedial course for science and technology, implemented in an agent based system. The final paper dealt with a travel support agent in Thailand, based on reinforcement learning.
In the panel, I made some remarks about the importance of interoperability in order to create critical mass, and the importance of critical mass in order to have impact on the way people learn. I encouraged participants to rely on existing standards in as far as they solve their problems, to identify problems that the standards do not solve and inform the standards bodies about those, and to develop their own solution for the problems
not “solved” by the standards. Wayne’s introduction focused on “getting small”, with content as well as services, so that the small components can be composed for a particular context. To illustrate how important getting small is, Wayne made the point that without interoperability in manufactoring, the industrial revolution would not have happened. Wayne also mentioned how important it is to “federate ourselves” and put ourselves together
around collaborative efforts. Nobuo Saito also emphasized the need to work well together. Bernd Simon talked about the evaluation of standards in terms of effectiveness, and the need for a business model, not only in financial terms,
but also as an organizational context, based on a clear understanding of the needs of the community they try to serve.
The questions for some reason were mostly addressed to me, somewhat to my embarassement.
- The first question related to the role of money. In fact, the question was really about the involvement of corporate players in developing tools and infrastructures. Dan made the point that there is substantial uptake, but that the corporate and academic world hardly speak to one another.
- The next question was about how LOM could be extended and further developed. I mentioned that the extensions to vocabularies, and the use of the classification category
- The last question was about how to create impact: I answered that we were not necessarily the best placed people in the world to figure out how to break into mainstream. Think of the WWW: it weren’t the folks at CERN or the W3C that developed the early versions of the Amazon, Google and eBay equivalent. (Of course, those were called Mosaic, Altavista, etc.)
Overall, there certainly was quite a bit of very relevant material presented and discussed, though the participation was rather tame and the day could have used some additional dynamic participations.