Metadata consistency: a major problem here is how to make a wikilike User Interface that allows to add constraints and stuff. Wiki users don’t like to get told “No, you can’t do that” — you would have to at least offer an explanation of why you can’t do that. So I guess this is a tough problem. Our solution? None — we just ignore anything that has more semantics than the subsumption relation, really.

KA bottleneck: we hope that people will enter it by hand. This is the advantage of the *many* people Wikipedia has. But you are right: we need to provide strong incentives. And we hope that the autogenerated lists really will provide such incentives. Would you rather annotate your country once and then create the country lists automatically, or update all the different country lists?

Reasoning: we don’t need reasoning for many of the advantages we are aiming for. Having a structured knowledge base can offer you a lot already. But the idea of reorganizing is — wow — really cool! Never thought about that. In that case, I guess, the reasoning happens offline, on your machine? So you tell your personal Wikipediabrowser “Hey, listen, this is my point of view on the world (aka ontology), now organize your stuff according to that”. Sounds very cool.

Argumentation: You are right :) And I need to talk to you about this one. Especially since the whole Semantic MediaWiki actually was inspired on argumentation-related work done in the SEKT project.

Hope to catch you tomorrow!
denny