[sc34wg3] Individual contribution on the U.S. N.B. positiononthe progress ion of Topic Map standards

Kal Ahmed sc34wg3@isotopicmaps.org
Wed, 31 Mar 2004 20:10:31 +0100

On Wed, 2004-03-31 at 19:28, Jan Algermissen wrote:
> Kal Ahmed wrote:
> > 
> > I detect a leading question, but I shall rise to the bait anyway - at
> > least partially :-)
> > 
> > I can load a topic map created by Ontopia into TM4J (and vice-versa) -
> > I can exchange information between multiple applications that implement
> > topic map processing as described by ISO 13250. I can describe existing
> > classification systems; thesaurii; CMS meta data and even relational
> > tables in terms of topic map constructs (as defined by ISO 13250).
> > 
> > That means I can exchange information between applications successfully.
> Hmm...so what? Of course you can since you use a common syntax. To me, this
> is data interchange, not information interchange. 

I disagree - this is not data interchange because a topic map allows you
to create a context for individual data. We are not talking about things
like "2", "31/03/2004" or the like, but topics which stand for subjects
and which have characteristics - thats information in my book.

> I think 'successfull
> information interchange" is achieved if I can hand you (and your application)
> my data *and* the semantics that govern the data (my "ontological commitments"
> if you like) and you would be able to understand them with the result that
> you would construct the exact same graph from the data that I do.

If I exchange topic maps, I have the rules for processing them to a
model (TMDM) and I know the means by which one expresses "ontological
commitments" - i.e. by topic characteristics. I don't need anything

> > 
> > There is much more that can be done based on existing ISO 13250 - it
> > needs time to be embedded in the information ecosystem and time for
> > research to be done and to push the boundaries of what the standard
> > provides. Why does *ISO 13250*  need to do more ?
> ISO 13250 does not enable what I described because it does not provide
> a mechanism to agree on semantics (what has been partly described
> by 'no ability to document merging rules').

You process the topic map as defined by the standard. Thats all there is
to it. I know what you are talking about and what you want to achieve
and I commend that goal - but its beyond topic maps as defined in ISO
13250. There is *nothing* in the standard with regards to "ontological
committment", or processing semantics or "ability to document merging
rules". So what you are asking for is out of scope with regards to a
restatement of the standard - it is something new, and it is not
something I need to be able to use topic maps as defined by ISO 13250.