> Chris Menzel:
> "Its only when we ground the semantics of the data in the behaviour of
> the application/organization that we have any hope of success, and then
> only after a long and painful process of testing."
> Sean Baker
> The point being that the terms of the definition need to be linked to
> real behaviour, and not to an interpretation of a model, since the
> objective of the exchange is always to get the organizations to work
> together, not to get the computers to agree that they could work
> together because they have compatible models.
While I agree philosophically with Sean, the practical reality reflects
Chris viewpoint. As a human, I think that it should be possible to speak of
one reality, as a computer scientist and engineers I see that it is not
possible - and in my oppinion never will be possible - to reflect one
reality in our systems, as all system are based on different
The problem is that our ambiguities start with our view of reality
(assuming there is really only one reality). In the moment we start to
perceive it, we already use a conceptualization of this reality. This
concepts do not reflect all aspects of reality, but only the necessary ones
to deal with the perception in the light of the task currently conducted or
the status the observer is in. If we use machines (or software in general)
to deal with this, we need a formal specification of this
conceptualization. If this formal specification concisely and unambiguously
defines concepts such that anyone interested in the specified domain can
consistently understand the concept’s meaning and its suitable use, we
normally call it an ontology.
In machines, the concepts used to formally specify the perception of
reality can differ in structure, scope, and resolution. This is already
hard enough for the static case, but if we broaden the view to the dynamic
case, they may differ in the information that is grouped together to make
sense (such as input parameters needed to start a process; these group of
parameters define a business object needed for the process, and the same
information can be used in different business objects, all with a different
context). In addition, even the same processes triggered by the same
business object can lead to different status changes in different machines.
This is the tested behavior of applications.
The problem is: two contradicting conceptualization can both be right ...
and that is the challenge to be solved by composable solutions.
Andreas Tolk, Ph.D.
Old Dominion University (06)
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (07)