[sc34wg3] Subject Schizophrenia

Robert Barta sc34wg3@isotopicmaps.org
Thu, 26 Aug 2004 15:58:16 +1000


On Wed, Aug 25, 2004 at 04:55:45PM +0200, Lars Marius Garshol wrote:
> * Martin Bryan
> | 
> | The purpose of facets in 13250 was, in my opinion, to encourage
> | subject schizophrenia. At certain points of time you are only
> | interested in a certain subset of the associations between subjects
> | - otherwise your map becomes too cluttered. I want a way of saying
> | "show me the subset of associations which conform to this facet of
> | the world view". If proxification is used in place of representation
> | as the process controlling bernersleeification , can I still do
> | this?
..
> I'd say that the main ways to do what you suggest are scope and
> annotations on the association types.

Martin,

Maybe I can answer this as we had to solve EXACTLY this problem a
while back.

Yes, one way is to tag the things of interest for a particular
environment (audience) with scope or use explicit associations.  In
our case this was not possible as the number and nature of these
environments are completely unknown at authoring time.

In our case we developed maps for lecturing purposes, say one about
XML, and in a particular course in a particular lecture we show
particular portions of that map. Topic after topic and for each topic
only a subset of the information disguised as a slide show. Compare,
for instance,

   http://topicmaps.it.bond.edu.au/mda/markup/xml/family/xml-family  [ raw topic view ]

with

   http://topicmaps.it.bond.edu.au/mda/markup/xml/family/xml-family@/users/rho/InTechII/   [ with glasses on ]

This is the same base content, but the latter "viewed in a particular
context".

While we had to write dedicated code for this last year, now we are
looking to completely develop this with TMQL. It can be done. And it
can be done quite elegantly.

\rho