Hello all,
thanks everyone for this very fruitful discussion. I have to admit that I
am still spoiled by practical applications driven by the need to integrate
existing/legacy systems and components into system-of-systems. This is what
drives my research in this domain. So far - and many practitioners still
believe so - the world was easy as everything was define by the data being
exchanged via interfaces. The components was a black box. However, in
particular when you use model-based components (where the model is a
purposeful abstraction of reality), the black box doesn't work. If you
connect boxes implementing different and not composable "abstractions of
reality," the stuff doesn't work - or even worse: it works technically, but
the results are wrong. Therefore, we looked for something to explain not
only what is good for the technical integration of systems
(integratability) and the alignment of the systems and the implementation
(interoperability), but we target the conceptual level, the underlying
models, constraints, and assumptions (composability).
Therefore, I fully agree with John Sowa's definition of what we need: (01)
1. A base logic, whose syntactic details are irrelevant. At the logic
level, RDF, OWL, SQL, and any of the formalized versions of UML are all
subsets of Common Logic. The very important practical differences between
them are irrelevant.
2. A collection of axioms that define all of the nonlogical predicates.
And by nonlogical, that means everything except the base operators of and,
or, not, some, every, etc. Even arithmetic and set theory are part of the
ontology. (02)
However, being spoiled and driven by the need to apply the ideas, I really
would like to add that ontologies and there defining needs comprise one
more thing: (03)
3. Ontologies are machine understandable. (04)
It should be pointed out that there is a difference between parsable and
understandable. I can also be argued that 3 is included in 1 and 2
implicitly, but having dealt with too many people who see ontologies only
as nice protege pictures on the one side (or other who declare to use
ontologies because the convert IDEF1X data into OWL data), I would like to
see this practical aspect of ontologies emphasized. Where I want to go with
this was summarized in my Keynote to the 20th ACM/SCS/IEEE Principles of
Advanced Distributed Simulation: What comes after the Semantic Web. (05)
Hope this contribution is perceived to be useful for the discussions,
Godspeed and 5 mph extra,
Andreas
============================== ;-)
Andreas Tolk, Ph.D.
atolk@xxxxxxx
Associate Professor/Engineering Management & Systems Engineering
242B Kaufman Hall
Old Dominion University
Norfolk, VA 23529
(757) 683-4500 (voice)
(757) 683-5640 (fax) (06)
Senior Research Scientist/Virginia Modeling, Analysis & Simulation Center
7000 College Drive, 111F Manning Bldg
Suffolk, VA 23435
(757) 686 - 6203 (voice)
(757) 686 - 6214 (fax) (07)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2007/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2007
Community Portal: http://ontolog.cim3.net/ (08)
|