John, (01)
Concerning: (02)
>In the field of artificial intelligence, it was taken for granted
>for many, many years that knowledge representation must always be
>done for some purpose (i.e., the pragmatics). Unfortunately,
>many people who are working on ontology have lost sight of the
>purpose -- or worse, some even think it is "unscientific" to
>talk about purpose, since that is a "subjective" term. (03)
Many of us try to strictly adhere to a methodology when developing
domain ontologies that focuses on the purpose of the ontology, i.e.,
the set of use cases, scenarios, competency questions for which it is
being built. I.e., stuff that drives out the requirements. The use of
competency questions goes back to Gruninger and Fox, circa 1994/5, and
then subsequently Gruninger-Uschold. When you can answer those
questions reasonably and at the right level of granularity as judged by
a domain expert, you're done. This means also that you can use those
questions/queries for testing. (04)
This methodology does apply to upper/middle ontologies, but of course
there are other issues that come into play. E.g., more complex
integration/interoperability, etc. (05)
Like Methontology-OntoClean, this kind of methodology extends software
lifecycle principles to ontology development. Sometimes the overall
subject is called ontology lifecycle management. Bill Andersen et al
have done work on this too. I am working currently on a chapter on
ontology architecture that addresses some of these issues. (06)
In a more disciplined world, the competency questions would be
associated with the given ontology, to facilitate reuse. (07)
Thanks,
Leo (08)
_________________________________________________________________
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 (09)
|