AW: [topicmapmail] Inference rules

[Murray Altheim]

Stefan Henke wrote:
> Hi Murray, Eric,
> thanks for your detailed replies. I have to agree to Eric. For my
> understanding, inference rules are an important feature to create new
> "knowledge" inside a topicmap. Murray, you are right that it is possible to
> express these inference rules in topic maps as topics and associations. But
> I think it is not a good idea due to reasons of constistence. Imagine the
> example of Eric using the cousin rule. If you model this rule as an
> association, each time a new person is added as topic, this association has
> to be added by the user that added the person topic. So, if it is not added,
> the map would not be in a consistent state. By defining it as a rule, it
> would be no problem as this rule is valid for all topics.
> Maybe I´m getting something wrong or I´m in a totally wrong direction, but I
> think inferencing is a very powerfull tool to "create" new knowledge.
> Eric, you mentioned TMCL in the context of inference rules. I´m not familiar
> with TMCL, do you think it is possible to define such rules with it?
> Regards,
> Stefan

Stefan,

I think you perhaps misunderstand what I've written. AFAIK, Eric
and I are not in disagreement. We both will be using Topic Map
topics and associations as the fundamental structures to do our
modeling; that's what's available in Topic Maps. But topics and
associations are at a fairly low level of modeling. As I'm sure
that advocates of the TMRM would assert, the concept of an
assertion is a structure that contains both the semantics and
the constraints of a specific statement (of logic or otherwise).
These assertions would be expressed as <topic> and <association>
elements in XTM syntax. A Topic Map engine that was unaware of
the specific semantics of an assertion expressed thusly would
not be able to reason upon it, but any inference engine that
accepted an input syntax of XTM using PSIs expressing the set of
known semantics could treat that XTM as any other form of
expression of the same semantics/statements.

And I did not mean to state that inferencing wasn't important. I
happen to be using a lot of inferencing in my own application.
All I was trying to state is that inferencing is everywhere. It's
so important that it doesn't need to be emphasized. Every SQL
database does inferencing, every interpretation of an XPath
statement too. It's just that applications of logic are so wide-
spread within computing that perhaps you and I are not talking
about the same thing.

Put it this way: I believe strongly enough in Topic Maps as a
foundation for knowledge representation that I started a Ph.D.
program to explore the possibilities. If you've been following
along with the conversation involving Jack Park, Dan Corwin,
and myself, you'd see that we've discussing inferencing at a
fairly high level, that of Conceptual Graphs.

As for TMCL, inferencing involves both logical structures and
constraints to check whether specific expressions fit within
those structures. So while there are certainly things that can
be done without the Topic Map Constraint Language, having it as
another tool in our bags will surely improve matters.

Murray

......................................................................
Murray Altheim                    http://kmi.open.ac.uk/people/murray/
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK               .

   The Resume of George W. Bush
   http://www.buzzflash.com/contributors/03/04/23_resume.html

                    The sleep of reason brings forth monsters -- Goya