Re: client-side image maps

From: Rosemary Michelle Simpson <rms__AT__cs.brown.edu>
Date: Thu Apr 20 2006 - 15:02:00 EDT

Okay, let me see if I can clarify, a useful exercise for me in any case!

Imagine classic pre-Web node-link directed-graph hypertexts in which the
node represents a piece of information and the edge, or link, represents a
relationship between two pieces of information. In those days, hypertext
was often explained by asking readers to imagine an encyclopedia with see
also links between the various articles. The hypertext was an electronic
version of the domain in which each article was a node and the see-alsos
were the edges, or links. Thus, it was possible to have a meta-level view
of a whole domain. In knowledge representation theory these were called
semantic nets - semantic net pioneer John Sowa's long and excellent
overview is at: http://www.jfsowa.com/pubs/semnet.htm - and an early
dynamic version could be seen in Kim Fairchild's SemNet, which was demoed
at SigCHI in 1986.

Image maps can be used as static representations of a semantic net, but
they are exactly that - static. I want a much more dynamic and
interactive representation that (1) dynamically generates the maps based
on a user's search requests, and (2) can be rotated and changed to reflect
the user's browsing and changing points of view. Visual Thesaurus
provides one solution, but quite frankly I find it enormously annoying to
work with in all the applications I've seen of the software and thus am
not pursuing that direction.

Does this help?

r

On Thu, 20 Apr 2006, Mark Anderson wrote:

> On 20/4/06 19:21, "Rosemary Michelle Simpson" <rms__AT__cs.brown.edu> wrote:
>
>> Thanks, Mark! Yes, that helps a lot. However, after writing you, I
>> thought about the matter further and realized that what I really want is
>> to create a dynamic semantic net web page(s) that accesses a Tinderbox
>> file, instead of a MySQL file, to dynamically generate a node-link
>> semantic net, one that changes its appearance as the user access it. The
>> user should be able to rotate it, zoom in on it, and ideally have a
>> focus+context view over it.
>>
>> I know that I could create a Java applet to do the 3D rotation, but I want
>> a dynamically generated webpage based on user access of the database. I
>> hate relational databases, and would like to be able to use Tinderbox
>> instead of the PHP-MySQL combination that has been suggested to me.
>>
>> Is this completely out of the question?
>>
>
> I only half understand what you want. To break things down a bit, what
> information are you seeking to get from tinderbox and in what sort of pieces
> - text or attributes. Alternatively is there something similar to what you
> want online that we could look at to help understand.
>
> In case it's not clear, the technique as before could be any code and not
> necessarily HTML.
>
> ---
> Regards
>
> Mark Anderson
> Tinderbox wiki gardener
>
> Tinderbox Wiki: http://www.eastgate.com/wiki2/wiki.cgi?TinderboxWiki
> aTbRef: http://www.acrobatfaq.com/tbx/index.html
>
Received on Thu Apr 20 15:01:59 2006

This archive was generated by hypermail 2.1.8 : Mon Jun 19 2006 - 13:29:09 EDT