[sc34wg3] N0391-0394: New SAM/XTM documents

Robert Barta sc34wg3@isotopicmaps.org
Fri, 18 Apr 2003 22:14:21 +1000

On Thu, Apr 17, 2003 at 01:25:19PM -0400, Patrick Durusau wrote:
> >Does TMM have a formalism for 'subsumption' and 'schema reasoning
> >consistency'?  Does it support reasoning (validation, not proof) at
> >all? What complexity does the underlying logic have? Is it PSPACE?
> >Without a minimal support of validation of schemas and theory behind
> >query engines,RDF/OWL is a too strong competitor, especially in the US
> >market, I would assume.
> > 
> >
> The TMM supports the building of TMAs, which is where you would find 
> support for 'subsumption' and 'schema reasoning consistence.' 


Sure. But why there? This is a perfectly valid while arbitrary
decision. If my goal would be to create a generic infrastructure for
Topic Map based information and I plan, of course, to allow queries
against the infrastructure, then my choices might be (and will be)
totally different.

Let's make an example: We develop an abstract, relational database
model. As abstraction we first choose tables, so all the data is in a
tabular form. This, by itself is worthless if we do not define (a)
update and (b) retrieval functionality. It is actually (a) and (b)
which convey the semantics of the data structure. The fact that these
are tables with bells and whistles is completely irrelevant from a
conceptual point of view.

This is why RDBMS are based on "relation algebra" (or, more recent on
the relational calculus). They define the operations. They constrain
the abstract data type.

> Can you 
> say a bit more about what would be required to support either of those? 

Phew, that is no easy question. I actually completely lack the vision
how resoning (again in the sense of semantic web) can be imposed
aposteriori to TMM or SAM. I know that SAM was not built for it. And I
assume the same for TMM. That's exactly my point.

For the TMCL/TMQL part it will depend a bit on the nature of the
agreed language how to map this onto SAM/TMM.

> >I seriously doubt that the (strategic) market will hold a place for
> >TMs much longer than a few months if we cannot credibly signal
> >progress.

> You may be entirely correct or entirely incorrect but other than just 
> hearing the words, I have no basis for evaluating your statements.

Well, ... yes :-). Talking to companies I got the impression that they
are not in the mood for interesting experiments. Times are tough if
you are not in a monopoly situation.

> >We have the big advantage as the 'second-comer'. Let's not throw this
> >away.
> > 
> >
> Glad to hear we have a "big advantage" but not sure from your post what 
> that is or how to capitalize on it. Can you say a few more words about this?

I also got the (possibly wrong) impression that we can profit a lot
from the RDF efforts and from the noise they make in the SW arena. It
is not always good to be the first in a market. You have to do things
step-by-step.  A second-comer (Microsoft is famous for that policy)
can clean things up. Maybe.