Re: FLOWS and URI

On Thu, 15 Sep 2005, Bijan Parsia wrote:

> On Sep 15, 2005, at 1:38 PM, Sheila McIlraith wrote:
>
> >
> > Bijan,
> >
> > I'm not sure I understand your question, but let me respond and if
> > I haven't answered your question, please let me know!
> >
> > The objective of FLOWS was to develop a FOL ontology describing
> > Web services.  We did it (or at least we did a reasonable first pass
> > at it.  There are lots of things we'd like to expand upon.)  FLOWS
> > is an FOL ontology of Web services.
>
> I completely understand that :)
>
> [snip]
> > OWL (and thus OWL-S) ontologies are Web documents and thus can be
> > referenced by means of a URI.  As such classes, properties, individuals
> > in OWL ontologies are all associated with URIs.
> >
> > If we wish to make FLOWS or a Web-variant of FLOWS into a Web document
> > (and there are lots of reasons for doing this) then it would make
> > sense to define concepts using URIs.  Nevertheless, that was not
> > our primary objective.
>
> You published a web document about it.

A web document *about* it is the key differentiator.   The ontology
is *not* a Web ontology.

> You intend for it to be used to
> describe web services. I believe you intend it to be used to put
> descriptions on the web.

That is a very reasonable trajectory, but that was not the point
of FLOWS.  We envisioned it as a Web-variant of FLOWS.

As I said above, that was not our primary objective with FLOWS.
If when we define a Web-variant of FLOWS (and there are lots of good
reasons for doing this) then it is obvious that we would want to
associate URIs with concepts as is the "semantic Web way".

So...why have an ontology of Web services that does not use URIs?
One (of many) answers:  For reasoning about Web services that are
not *semantic* web services as we know them now or for reasoning about
services offline.

You and I agree on the virtue of URIs and how they could be used in
a Web-variant of FLOWS.  I hope I have clarified the objective of FLOWS
that motivated the *choice* not to express concepts as URIs.

Sorry I won't have any more time to discuss this right now.

- Sheila






>
> Even without the latter, it's clear that you have a list of terms.
> Those terms could have associated URIs. This isn't *defining* the
> concepts using the URIs, it's just giving each concept a URI.
>
> For example, you have URIrefs for each section, e.g.,:
> 	http://www.daml.org/services/swsf/1.0/swso/#service-sec
>
> A simple naming scheme would have been to use
> 	http://www.daml.org/services/swsf/1.0/swso/#service
>
> To represent the concept, and content negotiation to get the HTML, or
> the KIF, or the CL, or the OWL, or the SWSL-Rules variant of the
> definition of that cocnept.
>
> In other words, I think the justification that the objective was other
> than to make it web accessible doesn't touch the reasonable point that
> in your document you could have provided URIs easily and naturally.
>
> In a sense, the complaint can be registered about your document, not
> your ontology per se.
>
> And there is a perfectly on point answer: We didn't get around to that.
> But given that 1) it's the *semantic web* services ontology and 2)
> there are *two* formalizations in proposed semantic web languages
> (SWSL-FOL and SWSL-Rules), it's sort of astonishing that you didn't
> have URIs!
>
> Maybe it isn't in the end important...but...it is weird.
>
> Furthermore, if I go to the overview:
> 	http://www.daml.org/services/swsf/1.0/overview/#roadmap
> """The complete axiomatization is given in first-order logic, using
> SWSL-FOL, with a model-theoretic semantics that specifies the precise
> meaning of the concepts. """
>
> """More specifically, SWSL is a general-purpose logical language, with
> certain features to make it usable with the basic languages and
> infrastructure of the Web. These features include URIs, integration of
> XML built-in types, and XML-compatible namespace and import
> mechanisms."""
>
> Now, I understand that SWSL and SWSO were in development concurrently,
> but the axioms *are* encoded in both SWSL-FOL and SWSL-Rules.
>
> Cheers,
> Bijan.
>
> P.S. I do like FLOWS :)
>
>

Received on Thursday, 15 September 2005 18:38:47 UTC