Thanks!
Looking at the picture on p. 8, I think there's a better way to
present that data. It's not clear at all that certain words in the
image belong to certain delineated regions. This is future work, but
maybe it would be better if there were lines of demarcation indicating
the grid layout of the regions. Or you could color all cells with some
(respectively differently colored) translucent overlay and the words
associated just pop up in descending order of importance (in a stack
in the radial direction, rather than a transverse stack) if you hover
over or click on it. I think they solved the combinatorial problem
that states you can color any map with no more than 5 colors or
something in the 70s. Extend that to our case (or not, which is much
simpler anyway because there are always 4 and only 4 adjacent
regions). I think that will be visually clearer.
On Sep 10, 9:41 am, Jason Baldridge <
jbald...@mail.utexas.edu> wrote:
> In case any of you are curious, you can check out the slides I'm using for
> my short presentation<
http://groups.google.com/group/textgrounder-open/web/ECADT-baldridge.pdf>at
> the
> TACC event today<
http://www.tacc.utexas.edu/education/humanities/emerging-communities-...>.