[Top] [All Lists]

Re: [ontolog-forum] Relating and Reconciling Ontologies

To: Barry Smith <phismith@xxxxxxxxxxx>
Cc: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Thu, 21 Apr 2011 15:12:48 -0400
Message-id: <4DB081B0.1020108@xxxxxxxxxxx>
Barry, Azamat, Ron, and Cory,    (01)

On 4/21/2011 10:51 AM, Barry Smith wrote:
> John,
> You are not responding to my request:
>>> Can you point to examples of practically useful mappings created
>>> and updated automatically through appeal to some sort of Lindenbaum
>>> lattice-based technology?    (02)

Lattices were used to organize the terms and definitions of
terminologies and ontologies long before anybody used the word
'ontology' outside of philosophy.  See the proceedings of the 1961
International Conference on Machine Translation of Languages and
Applied Language Analysis:    (03)

    http://www.mt-archive.info/NPL-1961-TOC.htm    (04)

CLRU was a hotbed of activity in novel ways of using lattices.
See the articles at that conference by Parker-Rhodes (p. 26)
and Margaret Masterman (p. 438).    (05)

More recently, the work on FCA (Formal Concept Analysis) has been
widely used for automatically developing, organizing, and relating
lattices of "concepts" (but they could just as well be called 'types',
which is my preferred term).  People have used FCA to verify, extend,
and relate ontologies written in OWL.  Type "FCA OWL ontology" to
Google for references.    (06)

To see little lattices generated from WordNet by the FCA methods,
type any word (for example, 'explore') to    (07)

    http://www.ketlab.org.uk/wordnet.html    (08)

Then type the same word to see the lattice generated from Roget's:    (09)

    http://www.ketlab.org.uk/roget.html    (010)

In the following slides (which I cited in my previous note), see
slides 61 to 81 for other methods based on the Lindenbaum lattice:    (011)

    http://www.jfsowa.com/talks/iss.pdf    (012)

And as I emphasized, you don't need to implement a theory in
a computer program in order to use it.  Every successful method
of aligning and revising ontologies conforms to the conditions
and constraints of the Lindenbaum lattice.  Knowing a little bit
of theory can help stimulate new ideas for algorithms that "use"
the theory, but don't literally "implement" it.    (013)

> I am convinced that to obtain the General Semantic Interoperability
> standard, costing hundreds billions per year, means to develop
> a single world reference model, in the first place.    (014)

> Just not possible. Who gets to decide? There are too many stakeholders.
> Each stakeholder will have trouble giving up a view of the universe that
> has served their organization for years in order to fix someone else's
> problem with this view.    (015)

> I don't think we will achieve this with a "single world reference model".
> Such a universal model would require the conception and integration
> of too many viewpoints and theories...
> Instead consider a "multi hub" approach where we attempt to minimize the
> number of reference models but accept that there may be more than one,
> even for a single domain.  Endpoints (the viewpoint specific schema and
> ontologies) may be grounded in more than one such reference models and
> reference models may be partially federated - this provides for
> federation where there is ANY reference ontology in common.    (016)

I sympathize with Azamat's hopes, but I agree with Ron and Cory.  I
don't believe that an ideal ontology of everything is possible until
science is finished:  i.e., all possible research questions have been
answered.  I'm not holding my breath.    (017)

John    (018)

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    (019)

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