ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Foundation ontology, CYC, and Mapping

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
Cc: Gian Piero Zarri <gian-piero.zarri@xxxxxxxx>
From: Gian Piero Zarri <zarri@xxxxxxx>
Date: Fri, 05 Feb 2010 10:37:47 +0100
Message-id: <4B6BE6EB.7020807@xxxxxxx>
John,

To be concretely useful, a system of primitives must be a) simple; b) general. This rule out both Pat's thousands of primitives and Wierzbicka primitives, which are too "surface level"  (see terms like "I", "You", "People", "For some time" etc.). In NKRL (Narrative Knowledge Representation Language) we have only 14 primitives, seven "semantic predicates" (Behave, Exist, Experience, Move, Own, Produce, Receive) and seven "functional roles" (subject, object, source, beneficiary, modality, topic, context). Terms pertaining to generic ontologies - we have our own ontology in NKRL, including presently more then 7,000 terms, but we have successfully used other ontologies in a standard Protégé style - are associated with the predicates through the functional roles to build up conceptual structures like, in simplified notation, (MOVE (SUBJ John) (OBJ BOOK_1) (BENF Mary)). Rules for setting up well-formed expressions of the previous type are specified by about 150 templates (axioms, if you want) that supply all the details of how to write a "transfer of a physical object" as in the previous example. Additional "sub-languages" and rules specify how to construct complex arguments of the predicate (a precious book with a red cover, Mary who is the sister-in-law of Paul), specify the locations, the temporal details etc. A powerful inference system, including causal reasoning, is used to take advantage from the advanced characteristics of the language. See http://www.springer.com/computer/artificial/book/978-1-84800-077-3.

Note that NKRL must be used to describe "narratives" (complex events) and to made inferences about them and it is not, then, an universal knowledge representation system like Conceptual Graphs or, to go back in time, Schank, Lenat or Ceccato's systems. The narrative domain is, however, sufficiently vast and important to justify the set up of a specific language to deal with it. My last NKRL application has concerned dealing in an 'intelligent" way with the "storyboards" built up in a gas/oil industry context...

Regards,


Gian Piero ZARRI

 

John F. Sowa a écrit :
Pat and Chris M,

I have read the following slides, which make some interesting points
and cite some useful references:

http://www.micra.com/COSMO/TheFoundationOntologyForInteroperability.ppt

First of all, I have a very high regard for the work by Anna Wierzbicka
which I have been following for nearly 30 years.  (I cited her early
_Lingua Mentalis_ in my 1984 book.)  I also agree with Cliff Goddard
that the arguments against that kind of research are *bad*.

But I must emphasize that the so-called "primitives" that Anna W. and
others have proposed are most definitely *not* primitives in the sense
that mathematicians use.

Anna W's primitives and the Longman's primitives may be useful as
rough guidelines in a methodology for designing good human-factored
representations.  I suspect that they could support a better basis
for pedagogy than set theory, which turned out to be a disaster
when it was inflicted upon innocent children by inept teachers.
I would encourage such research.

But Anna W's "primitives" are extremely vague and squishy.  She manages
to use them to "define" lots of different terms across many different
languages and cultures.  I find those exercises intriguing.  But her
definitions are so vague that they would be totally worthless for
formal ontology.  The give *ZERO* evidence in support of Pat's claim
that a precise FO is possible or that it would have the slightest
value for interoperability.

Furthermore, the FO that Pat has suggested is a closed system.
Following is the point I made, which Pat did not address in the
more recent note:

JFS> I suspect that what you have in mind is a closed system:  a
 > finite set of relations (i.e. primitives) that are fully specified
 > by necessary and sufficient conditions...
 >
 > Then all possible terms that could be defined in your system would
 > consist of closed-form definitions of the following kind:
 >
 >   Every term T is either a primitive or it is defined as synonymous
 >   to an _expression_ that is composed only of primitives.

A closed system is a dead end.  It rules out most of mathematics,
including systems that are important for physics and engineering.

Since it is very difficult to analyze a system with 2148 "primitives",
I used the many versions of set theory as an example.  Chris raised
the following point, which I agree with:

JFS>> But Chris M. used the example of set theory, all versions of
 >> which have the same two "primitives" -- subsetOf and elementOf.

CM> elementOf, i.e., ∈, is the only primitive you need (and the only
 > one you'll find in most modern texts).

I agree.  I should have added a qualification to avoid that objection,
but I'd like to generalize the issue in a way that makes my argument
stronger.

The reason why I claimed two "primitives" for set theory is that
I was thinking of a broader range of theories, which would include
the many versions of mereology and set theory as special cases.

Some versions of mereology deal with continuous stuff, which is
infinitely divisible (no atoms).  Other versions have atoms, but
no continuous stuff (sometimes called 'gunk').  And other versions
of mereology contain both atoms and gunk.

It is possible to treat set theories as special cases of mereology
in which there is no gunk.  In that case, the subsetOf operator of
set theory is a relabeling of the partOf operator of mereology,
and there is the more basic elementOf operator, which can be used
to define subsetOf.

When you consider the many versions of *both* mereology and set
theory, you get a huge number of ways of formalizing the vague
notions of parts, wholes, and collections -- and all of them are
based on just one or two "primitives".

You can also look at all the many axiomatizations of the notions
of possibility, necessity, and other modal operators.  Those are
further examples of the open-ended number of ways of formalizing
the vague modal terms that are common in NLs.

I'm sure that every one of Anna W's "primitives" and any others
that might be proposed by Goddard and others can be similarly
formalized in an open-ended number of ways.  Picking just one
of those formalizations over any others would be purely
arbitrary.

Conclusion:  Pat has not shown any credible evidence for the
claim that an FO based on such primitives would help support
interoperability among practical computer systems.  On the
contrary, its closed nature would probably create more
obstacles than it could eliminate.

John

 
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
 

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (01)

<Prev in Thread] Current Thread [Next in Thread>