Re: SUO: Re: Missing Ingredients
Murray,
I agree:
> ... I don't really need a SUO. What I'd like
> would be a means of aligning my work not with a SUO but with
> specific other ontologies in related domains....
That is, in fact, very close to what I have been advocating:
a modular system that supports a hierarchy of modules, which
can be used separately or in a mix & match combination for
various purposes.
> All I want is interop between domain ontologies, not interop
> with everything in creation. Authoring ontology to authoring
> ontology, medical ontology to medical ontology, without loss.
Yes. In fact, this approach would solve the engineering problem:
it would mean that people who really have problems to solve
in the job they are working on would develop an ontology for
something they understand instead of some abstract cream on top
of the ontology pie.
> Now, I know that there are good reasons for upper level work
> to be done, with the idea that the SILOs need to align their
> things with upper things, but I'm not sure I believe that it
> is possible to create a truly general purpose upper ontology,
> and if it is possible, we'd still need that set of documents
> I mentioned previously; we need the Lego building blocks for
> building *any* ontology, ROSO, ULTO, SILO or SUO.
I believe that the top level should be very small. Something
much closer to the KR ontology of my 2000 book than something
as large as SUMO or OpenCyc.
> ... That's one
> of the reasons why I think Pat wrote LBase....
Actually not. LBase is logic, not ontology. LBase is
actually an application of the Common Logic model theory
to OWL and RDF.
John Sowa