Re: [semanticweb] how to explain to humans the term ontology

Thanks Azamat for picking up my question about 'one ontology = one view 
of the world'
I have had the perception over the years  that people have been trying 
to obtain an unified view of the world
:-)

>
>
> Knowingly or unknowingly, we are all after a unified framework 
> ontology (UFO) integrating upper-level ontologiesgeneral modelling 
> languages (as semantic web ontologies, UML, OO programming languages, 
> etc.) as well all the mutitude of domain-specific ontologies and 
> perspectives.

  This means a UFO does not contradict nor exclude the existance of a 
multitude of differnet domain ontologies, rather, it is based on it  (
right?)

In order to be 'valid'  I guess it will have to serve every possible 
purpose and be useable by every possible application
(I ll be intrested to see the validdity of a UFOs tested )

> The history of all science is marked by the quest of most unifying 
> theories and models about the world and its parts, like a theory of 
> everything [physical] in theoretical physics.  But, unlike this, 
> Ontology is a formal theory of everything [physical, chemical, 
> biological, mental, social, cultural, or informational, as web 
> resources].


The problem with science, is that it only considers 'true' whatever it 
can understand'/proof., and that science itself is coming to terms
with its own limitations.

Science  does not model reality in its entirety either, but only the 
parts of reality that it knows,  the rest, whatever it cannot explain, 
it disregards, or at least, it does not take into account. (That 
includes so many natural and social phenomena)

A unified view of the world  includes, and is based on, all the (valid) 
views that exist. The validation of a ontology - proof of concept - its 
is usefulness to the
purpose for which it was created (correct me if I am wrong)

In systems development  the first  essential step system is 'determining 
the boundaries', and the systems interface with
other systems boundaries.  kind of ' what are we looking at' question.

Of course it would be nice to 'look at everything'  but that somehow may 
shift the system focus away from its functional goal.
So we make arbitrary distinctions, approximations and some compromises, 
to make our (applied) work 'feasibile'.
Here is maybe where theoretical and applied work have two different 
dimensions.

A boundary allows me to indenfity and hold true certain conditions 
within the system that I can refer to as 'axioms'. Without which, the 
system would not be able to
accomplish its goals

The ability of that system to work outside such boundaries, is what I 
think we are tyring to achieve with web services
:-)  and it implies the ability to transcend   boundaries - system 
taboos? - and still achieve its (expanded) goals.

This  may imply a shift in the initial goals:

When we  model reality in its depth and complexity, and look beyhond 
systems boundaries, we realise that everything in the known universe  is 
related to everything else, and that  a few  common elements constitute 
its substance , and a few universal axioms,  determine the behaviour of 
everything .

When it comes toa valid ontology, its pobably useful to be able to look 
at the world both ways: from an individual 'user' perspective, and from 
a global 'domain' perspective (the sum of all knowledge)

Whatever is likely to be 'true' in all  instances, may well be worth 
modelling.

Thats where science and philosophy merge.


Paola Di Maio

Received on Friday, 13 January 2006 03:17:20 UTC