Michiel Hildebrand wrote:
Designing user interfaces for "new" ways of exploration is indeed
difficult. Evaluating these interfaces is even more difficult.
In our current research we have deliberately simplified the tasks that
we try to support with our designs, so that we are able to do some
form of evaluation. (Term search from multiple thesauri to support
annotation).
The nice thing of your previous projects (timeline, exhibit) was that
they were evaluated by the community. Their success may indicate that
they are useful. Is this approach your intention? I haven't read your
thesis yet, maybe the answer to this question is in there.
Do you think a similar process can happen for parallax?
If so, what are the tasks people will perform using this interface?
I like your screencast as it gives a good idea of the functionality,
but the actual results of the "interactive query building" is not
something I would need. Who are the users that will need it? Have you
already had reactions of users that good find (create sets) that they
really needed?
It's very perceptive of you to ask about the tasks that Parallax is
presumed to address, and who the users are. I don't have a specific
answer beside "browsing graph of data more efficiently".
I tend to think that contemporary graph-based data browsers either fly
the user at 50,000 feet and show her the whole world in one window below
(render a huge data graph as a huge visual graph), or leave her at the
street level to wander around on foot (single resource view). I'm just
wishing to provide her a car. Perhaps the good thing is that the car
doesn't come with a destination built in. (It'd be quite bad in real
life if you need different cars to go grocery shopping and to go to
work, for example.)
As for user reactions, here is one:
http://www.spellboundblog.com/2008/08/16/freebase-parallax-search-olympic-games-facts/
I don't know if the results were useful to the blogger or she was just
having fun browsing around.
David
