Wednesday, May 9, 2007

WWW2007, workshop on the collaborative construction of structured knowledge

I just returned from the World Wide Web conference in Banff, Alberta, Canada and now I'm frantically trying to assemble some notes before it all fades into the memory fog.

The first day of the conference I attended a workshop on the collaborative construction of structured knowledge. The purpose of the workshop was to explore requirements, opportunities, and current tools for enabling multiple people to contribute to the same knowledge resource. The resources ranged from tags at the bottom, through wikis, to the creation of description logic ontologies. Below I offer a few highlights of the day from my perspective, starting with my own little brush with fame.

Prior to the workshop, there was a competition involving 6 new collaborative KR tools. In this competition, the users of the tools competed with one another based on how many bits of knowledge that they constructed with the tools and on the quality of the comments that they provided to the tool builders. The contest incentivized people to actually try the (mostly pre-alpha..) tools thoroughly, thus providing the tool builders with good feedback and stimulating discussion at the workshop. At the end of the workshop, Natasha Noy, one of the organizers and a pre-eminent figure in the world of ontology based knowledge representation, preceeded her presentation of the results of the competition by acknowledging that a paper by "Mark Wilkinson" provided some of the inspiration for the idea of the competition. Ahem.

As the first author of the paper that she was referring to, I was 99% thrilled to have it recognized by such a prestigious figure but (at least) 1% distressed that it was referred to as Mark's paper without any mention of me or any of the other author's (though the others had a very limited role). I spoke up and she immediately apologized - and did so again in person after the end of the session. I have no hard feelings about it, and like I said, am delighted that people like Natasha, who are at the very top of this field actually knew about my work; however, the whole episode brings up a lot of interesting facets of the culture and perhaps the economics of academia (and perhaps all human society in the networked world) that might help to shed some light on the questions about blogging I am still trying to work through. I'll talk about this more in another entry..

A few highlights:
1) Andrew Gibson - Described the need for people of different backgrounds and with different desires to interact with ontologies. This provided some of the motivation for a web2.0 classification system for ontologies. Each ontology would have its own 'profile' consisting of meta-data (kept distinct from the ontology file) that would include information about aspects such as authorship, revision state, history, purpose, deployments, users, and even threaded discussions.

Cool, lets make it happen.

2) Michael Backhaus presented BOWiki - a biology specific semantic media wiki. Interesting to see if the work they did in grounding it in an upper ontology of function will matter. (Only if they can actually get users..)

3) SOBOLEO - My personal favorite entry in the challenge because a) it worked for me b) it was simple and c) I enjoyed the live edit tracking feature.

4) Also from SOBOLEO group: A very clever idea called imagenotion. Basically, it replaces concept term labels with pictures. I think this is an obvious (in retrospect) and potentially extremely powerful thing to enable. Is a picture worth a thousand words? Has to be a paper title for them if it isn't already. The incorporation of multimedia in the ontology browsing/creation experience is going to be exciting because it will really make it much easier (I think) to communicate the concepts with real people (not robotic logicians).

5) Great keynote "Stone Soup", "freebase" on economic lessons for the now web.

More to come on this..