[Sigia-l] Information Visualization

[Listera]

"Karl Fast" wrote:
 
> But I *disagree* with the suggestion that (a) infoviz is a failure
> unless it can deal with billions, and (b) that solving a problem
> with a million records isn't really useful and represents trivial
> problems.

This is *not* what I said, at all. My points have been about the scalability
aspect of infoviz, though I specifically raised the non-technical aspects as
being of interest. If you want to restrict infoviz to the domain of
non-scalable problems, I can understand it. Is it useful? Not sure.

What's intellectually *and* commercially interesting are the datasets that
are not only large but also difficult to make sense of using
conventional/procedural means.

You mention elsewhere Amazon. Amazon.com has one the largest DBs in the
world at over 13 terabytes. The value of that DB is in its ability to
inter-relate multitude of data. I'm sure Bezos would love to find a way to
parse through BILLIONS of records, transactions, links, navigation patterns,
etc. and visualize what's been hidden behind the sheer volume of data. I'm
also sure he'd pay dearly if infoviz could make that happen and increase
Amazon revenues. 

So to me, the *commercial* success of infoviz is directly proportional to
its ability to deal with large datasets whose owners are, after all, in a
unique position to pay for its development.

Is infoviz relevant/useful for smaller problems? No doubt. Indeed, I suspect
that's where a lot of UI innovation may come from. I narrowly objected to
the notion that because you can solve retrieval issues on a small scale,
that somehow means that the same model/UI can through some technomagic apply
to large datasets. That's fundamentally flawed thinking. There are no
technical trends that foresee that in the near future that I'm aware of. To
claim otherwise is being irrational.

Ziya
Nullius in Verba