[Sigia-l] Information Visualization

[Karl Fast]

> The gist: infoviz doesn't work, by and large, because it's just not 
> usable, and it doesn't allow people to accomplish tasks in any
> better way.

Ahh, but what tasks are we talking about?

This is something Ziya brought up when he asked about "non-technical
barriers" to information visualization. Todd also touched on it in
his post about infoviz weaknesses.

I've been reading the infoviz literature. As part of this I have 
asked critical and skeptical questions, many of them similar to
those expressed on this list.

The literature seems to have little understanding of the
relationship between tasks, users, and visual representations.  The
field is dominated by computer scientists interested in building
neat-o tools. This makes for lots of gee-whiz applications and gets
you research money, but it's not getting at some of the underlying
fundamental issues (not yet). I say this more as observation than
damning criticism.

What are these issues? I can't give you an itemized list. Not yet.
I'm working towards that, but this research thing takes time and
my reading list is gigantic.

One of the biggest issues is the lack of decent frameworks to guide
the development of effective visualization applications. We don't
know enough about what constitutes an effective visual
representation, how best to interact with it, and how this all
supports the development of richer cognitive structures
(visualization is about more than just improving retrieval, a common
misconception)