[Top] [All Lists]

RE: [ontac-forum] Problems of ontology

To: "ONTAC-WG General Discussion" <ontac-forum@xxxxxxxxxxxxxx>, "ONTAC-WG General Discussion" <ontac-forum@xxxxxxxxxxxxxx>
From: "Peterson, Eric" <EPeterson@xxxxxxxxxxxxxxxxxxx>
Date: Thu, 18 May 2006 09:04:06 -0400
Message-id: <9BFFF3547B8B264891D4CF5C6E712512160B74@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx>

From: ontac-forum-bounces@xxxxxxxxxxxxxx on behalf of John F. Sowa
Sent: Wed 5/17/2006 6:57 PM
To: ONTAC-WG General Discussion
Subject: Re: [ontac-forum] Problems of ontology

Azamat, David, and Eric,


   EP> All my thoughts and effort toward mapping and merging upper
    > ontologies have been centered on the creation of a merged lattice
    > of types and and a merged lattice of relations.  For my purposes,
    > this is where I seem to be getting value.  Theory boundaries seem
    > to me to be more arbitrary and less useful stuff with which to
    > integrate databases.  I'm still not sure if we actually differ
    > on this point.  I'm suspecting that we agree.

JS> I agree that working with a lattice of types and relations is
> important.  But as soon as you start writing axioms for those
> types, contradictions begin to proliferate.  I like to quote
> Peirce's remark (CP 4.237):

Many of us (myself included) have spent the governments maney for years trying to put those axioms to practical use.  I'm not sure that we have shown any real value to the government yet.  Personally, I am currently getting much more job satisfaction out of working on the more ordinary task of laying the foundation suitable for axioms to adorn.  I find the humble foundation work to abound with consuming technical challenges.

So when I talk about working toward a merged upper model, I'm attempting to have a much simpler discussion than I believe that you are.
I claim that an ontological model suitable for merging databases needs no baroque axioms.  Where the need for fuller logic comes in, I claim, is in mapping the database models to the core ontology hub model.  And then only occasionally.


JS> The final sentence corroborates the point that Lenat makes about
> Cyc -- the most important inferences are in the microtheories,
> which are, in Peirce's terms, "very narrow subjects" about
> which it is possible "to be pretty precise and fairly certain".

Doug's engineer who wrote the OWL translation of Cyc, could have but didn't bother to translate or maintain the microtheory structure.  I haven't heard any complaints about that.  My claim is that microtheories are not needed for a database federating ontology.  At best microtheories are arbitrary engineering conventiences.  At worst they get in the way and bag up things that do not always belong together.  Microtheories do not form anything but flat lattices unless an engineer specifically and intentionally designs them to subsume on another.  Without intentional engineering, the likelihood of any two non-trivial microtheories subsuming one another staggers my poor imagination.

This is the problem that I hit into when I think about a lattice of theories.  Maybe this timo I will converge to your point of view.  I suspect I'm missing something, as I am known to do.



-Eric Peterson

Message Archives: http://colab.cim3.net/forum/ontac-forum/
To Post: mailto:ontac-forum@xxxxxxxxxxxxxx
Shared Files: http://colab.cim3.net/file/work/SICoP/ontac/
Community Wiki: 
http://colab.cim3.net/cgi-bin/wiki.pl?SICoP/OntologyTaxonomyCoordinatingWG    (01)
<Prev in Thread] Current Thread [Next in Thread>