Folks,
Like Leo, I'm wondering why there should be an apparent dichotomy in
these arguments. The "right" answer is really that both top-down and
bottom up are required. To construct ontologies without the benefit of
reusing top down, generic models of temporal and spatial information, set
and class theory etc is to create theories which are uniformed by best
practices. Concentrating on abstract theories alone results in ivory tower
theories that may be intellectually elegant but are unconnected with
solving real application needs.
Sadly, this dichotomy has been the state of the practice, in large
part. Academics have not often embraced the large scale needs of industry
and industrial folks have been too quick to just code up whatever is expedient.
I'm hoping that my group's approach of providing a free, formal upper
ontology, and incrementally providing free domain specific ontologies will
help, at least in a small way, to change the state of things. (01)
Adam (02)
At 10:15 PM 1/19/2003 -0500, Sam Hunting wrote:
>On Sun, 19 Jan 2003, Leo Obrst wrote:
>
> > Amen, Bill. I am a strong advocate of top-down ontologies (and so I recuse
> > myself), and so will not say more than you've said, except: The best
> approach (in
> > my view) is a mixed top/bottom: harvest the documents you need to cover
> from the
> > bottom-up, induce their requirements upward, and come down from the top
> with
> > principles that seem to match.
>
>Pragmatically, I'm with you, Leo.
>
>And I'm not trolling when I was the question --
>
>The phrase "incomplete ontological effort" intrigues me. What is the
>operational definition of a "complete ontological effort" -- surely this
>is entirely a function of level of effort?
>
> > "William E. McCarthy" wrote:
> >
> > > No matter how loud the marketing drumbeats get for UBL, there can be no
> > > disputing that it represents at its core a bottom-up approach to
> > > developing a business process ontology. You can generalize documents
> > > forever and still not unearth basic economic and business principles of
> > > economic exchanges. Documents add unnecessary components (like
> details of
> > > manual reconciliations/bookkeeping and/or paper exchange). Documents
> also
> > > hide components that should be analyzed and generalized (like
> > > implementation compromises of ideal models). Additionally, I do not see
> > > how the state of affairs at a process or collaboration level can be
> tracked
> > > by seeing which document was sent last.
> > >
> > > Harvesting existing knowledge (as UBL does) is a good partial
> > > approach. However, until UBL attacks some of these conceptual issues
> from
> > > the top down, it will be an incomplete ontological effort, no matter what
> > > personalities or companies are behind it.
> > >
> > > Bill McCarthy
> > > Michigan State
> > >
> > > moy rAt 02:05 PM 1/18/2003 -0800, you wrote:
> > > >As the UBL release is nearing, we are starting to see their story
> covered in
> > > >trade press now.
> > > >
> > > >Take a look at the eWeek article:
> > > >
> > > > Push for UBL Protocol Gathers Momentum
> > > > By Darryl K. Taft / January 13, 2003
> > > >
> > > > http://www.eweek.com/article2/0,3959,825911,00.asp
> > > >
> > > >-ppy
> > > >
> > > >--
> > > >To post messages mailto:ontolog@xxxxxxxxxxxxxxx
> > > >An archive of the [ontolog] forum can be found
> > > >at http://ontolog.cim3.org/forums/ontolog
> > >
> > > Bill McCarthy
> > > Michigan State University
> > > 517-432-2913
> > > http://www.msu.edu/user/mccarth4/
> > >
> > > --
> > > To post messages mailto:ontolog@xxxxxxxxxxxxxxx
> > > An archive of the [ontolog] forum can be found
> > > at http://ontolog.cim3.org/forums/ontolog
> >
> > --
> > _____________________________________________
> > Dr. Leo Obrst The MITRE Corporation
> > mailto:lobrst@xxxxxxxxx Intelligent Information Management/Exploitation
> > Voice: 703-883-6770 7515 Colshire Drive, M/S W640
> > Fax: 703-883-1379 McLean, VA 22102-7508, USA
> >
> >
> >
> > --
> > To post messages mailto:ontolog@xxxxxxxxxxxxxxx
> > An archive of the [ontolog] forum can be found
> > at http://ontolog.cim3.org/forums/ontolog
> >
>
>Sam Hunting
>eTopicality, Inc.
>
>---------------------------------------------------------------------------
>"Turn your searching experience into a finding experience."(tm)
>
>Topic map consulting and training: www.etopicality.com
>Free open source topic map tools: www.gooseworks.org
>
>XML Topic Maps: Creating and Using Topic Maps for the Web.
>Addison-Wesley, ISBN 0-201-74960-2.
>---------------------------------------------------------------------------
>
>--
>To post messages mailto:ontolog@xxxxxxxxxxxxxxx
>An archive of the [ontolog] forum can be found
>at http://ontolog.cim3.org/forums/ontolog (03)
--
To post messages mailto:ontolog@xxxxxxxxxxxxxxx
An archive of the [ontolog] forum can be found
at http://ontolog.cim3.org/forums/ontolog (04)
|