To address a couple of points people raised...
1. We'd love to work more closely with the Wikipedia community! It's a
great model of dissemination and gathering of knowledge. Certainly we'd love to
discuss whether there's any way our technology might be useful to you, or that
we could collaborate.
2. One person questioned our use of Wikipedia text
in our screenshots. We do link to Wikipedia (the site, the archival database,
and the Wikipedia entry on Wikipedia) but we could also either (a) alter the
text so it is unreadable; or (b) since readable text makes the diagram more
understandable: with Wikipedia's agreement, we could put a note on our gallery
page saying "Article text (c) 2003 Wikipedia, reproduced by permission." Sorry
for any confusion, this is a confusing area! (in fact, wasn't there recently a
whole thread on fair use on this mailing list?)
3. Another poster asked
about our diff method: I think the main difference between ours and yours is
that ours is operating at a finer granularity (roughly at the sentence level,
rather than the paragraph level). We'd be happy to talk in more detail, if you
want!
4. We have been finding some fascinating patterns about the different ways
in which people collaborate in Wikipedia. We are currently writing a paper about
this and would be happy to share it with the community after we are done.
Again, it is great to hear all the excitement about the project here and it
would be great to figure out ways to collaborate and hear more
feedback from you.
best,
Fernanda Viegas and Martin Wattenberg