[Top] [All Lists]

Re: [ontolog-forum] Looking forward at the past

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Sun, 19 Oct 2008 10:26:02 -0400
Message-id: <48FB437A.7060206@xxxxxxxxxxx>
Sean,    (01)

I certainly agree with that comment as well:    (02)

SB> The comment was that you probably need a bigger range of
 > fundamental relationships to cover what is available on the web.
 > Most web pages are presentations rendering a representation
 > (HTML, XML) describing a "property" of something at some point
 > in its lifecycle from a particular viewpoint. There is quite
 > a difference between buying a Ferrari on E-bay and buying a
 > picture of the Ferarri, so knowing whether the reference is
 > to the picture or the thing it presents is quite important.    (03)

My comments about these and many other issues have been scattered
among a large number of notes, and I'll have to gather them together
in a single proposal that spells out the details in better organized
web page.    (04)

SB> We might sensibly have a debate whether views, representations
 > etc are "as foundational" as subtypes or part/whole - though not
 > this week, as I'm travelling.    (05)

As I said in other notes, the term 'foundational ontology' is one
that Pat Cassidy suggested and I adopted because it is about as
good as any other and better than most.  But we definitely need
a detailed specification of what kinds of information belong in
what parts of the foundation.  Following is a very brief summary:    (06)

  1. The backbone of the foundation is more of a taxonomy than a
     detailed, fully specified ontology.  It is mainly based on
     two relations:  type/subtype and part/whole.    (07)

  2. For specific applications, more details must be stated in an
     open-ended library of fully axiomatized modules, whose type
     labels are organized as in point #1.    (08)

  3. The modules in #2 are related by a lattice, which shows the
     dependencies, compatibilities, and incompatibilities among them.    (09)

The larger range of relations you are asking for would be specified
in the library of modules.  Some of them are general enough to be
used and reused in a very large number of applications.  The general
modules that specify them would be located high up in the lattice.    (010)

You can think of the general modules as something along the lines
of what people have been calling an "upper ontology".  But I would
drop the requirement that any particular module must be an absolute
requirement.  Instead, a module is only "required" relative to
other, more specialized modules that depend on it (i.e., the lattice
specifies the dependencies among modules).    (011)

Some modules may be so specialized that they can only be used in a
very narrow range of applications.  But even a specialized module
may have generalizations higher up in the lattice that can support
interoperability with a larger number of applications at a less
detailed level.    (012)

For example, a calendar program may use times and dates with
general axioms that are common to almost every module that uses
times and dates.  But the detailed axioms needed for communication
with the Mars rover may need very specialized details that go far
beyond what a calendar program needs.  However, the people working
on the Mars rover can use the general axioms of a calendar program
when setting up a meeting with other people on earth.    (013)

John    (014)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (015)

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