JoeSinkwitz - 2:00 pm on May 11, 2007 (gmt 0)

[Cygnus, that would show you first-order co-occurrence, but would there be enough data in that kind of limited set to be able to include how the results have been influenced by second-order co-occurrence?]

No, it probably wouldn't be enough to explain everything that we're seeing thusfar, but it'd probably help to develop a primer for one's niche. In terms of the second order, third order, etc...keep running it periodically, making note of new sites that pop into each section and how the co-occurence of the previous sites change -- that is obviously going to be a bit more tricky and I haven't really thought about ways to script it just yet. However, if you can build enough data that shows what the first order co-occurence is, and see how sites move into that EOS issue on the co-occurence changes, the changes themselves might be enough to hilight how the themes move (and possibly collide).

The amount of data required to do it right would be massive, but a hack and slash method of determining the moving targets of how the co-occurrence evolves for themes, using limited data, is like a using landmarks as a navigation system...good enough in most cases.

Hopefully someone else has a more elegant solution that what I'm deriving with my diet coke fueled mind.

Cygnus