W3C home > Mailing lists > Public > www-rdf-logic@w3.org > October 2000

Re: basic decisions underlying DAML-ONT

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Thu, 12 Oct 2000 13:11:07 -0400
To: jhendler@darpa.mil
Cc: www-rdf-logic@w3.org
Message-Id: <20001012131107M.pfps@research.bell-labs.com>
From: Jim Hendler <jhendler@darpa.mil>
Subject: Re: basic decisions underlying DAML-ONT
Date: Thu, 12 Oct 2000 11:24:01 -0400

>   But here's the problem - remember the new game! We're on the web, so 
> someone somewhere defines something as a printer, some catalog 
> somewhere defines its cost as $1000.  Is the system inconsistent if 
> we don't return that printer as an expensive printer???  That is, can 
> we insist that there must exist a distributed mechanism that will 
> somehow find these two facts (which are likely on different pages, 
> maybe even pointing to different name spaces that in turn point to 
> other things that eventually both share the same DefinedClass)??

My observation has nothing to do with what inferences are required.  You
can take a very powerful logic and build a partial reasoner for that logic.
As long are you are (very) careful, you can use that reasoner to perform
useful tasks over statements in the logic.

However, to me it seems very silly to produce a language that allows one to
create ``defined'' classes in a very complicated manner, but doesn't allow
one to create them in a simple manner.  Yes, if one expected people to type
in the class definitions then this would make ``defined'' classes much less
prevalent.  However, I expect (and I think that this is a common
expectation) that many of the class definitions will be created via some
sort of ``easy interface''.  It is certainly possible to build an interface
that would turn a condition (such as individuals with a rich friend or some
boolean combination of DAML restrictions and qualifications) into
a set of DAML statements that create a class with that definition (such as
in my original message).  If this can be done, then why not include a nicer
way of stating this important modelling idiom.

>   The idea of doing logic and inference in a widely distributed, 
> uncontrolled, dynamic, web-grounded way is new, and challenges a lot 
> of basic assumptions.   A real challenge is this notion of a 
> declarative language that is separate from the inference engine(s) 
> that use it.  And, if you'll pardon a personal observation for a 
> moment, that is what makes it so exciting and so much fun -- many 
> things we've taken for granted in KR for many years are suddenly 
> called into question.

The separation between the declarative language and the inference engine(s)
is actually rather old.  The line of development from KLONE to KL-ONE to
NIKL to LOOM to POWERLOOM lives in this arena.  One could say that all
first-order reasoners live in this arena whether they are partially
complete or not.  I suppose that this point could even be stretched to
complete description logic reasoners, as there are multiple inference
engines available for many description logics.  These engines are even
qualitatively distinguishable based on behaviour.  (For example, DLP
quickly solves some formulae that KSAT cannot effectively solve, and vice
versa.)

Peter Patel-Schneider
Received on Thursday, 12 October 2000 13:12:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:37 GMT