W3C home > Mailing lists > Public > public-sws-ig@w3.org > November 2005

Re: [fwd] Draft charters for work on Semantics for WS

From: David Martin <martin@AI.SRI.COM>
Date: Tue, 29 Nov 2005 18:03:06 -0800
Message-ID: <438D085A.9030204@ai.sri.com>
To: Joel Farrell <joelf@us.ibm.com>
CC: "public-sws-ig@w3.org" <public-sws-ig@w3.org>

Joel Farrell wrote:

> One point of clarification.  There is a difference between the annotations
> in WSDL-S and those in described in the of OWL-S WSDL grounding.  In WSDL-S
> a modelReference or a SchemaMapping refer directly to a domain ontology.
> The OWL-S WSDL extensions refer to parts of an OWL-S ontology, which in
> turn can refer to a domain ontology.

That's an interesting point, but I don't think the distinction holds up 
across the board.  It's a straightforward distinction with respect to 
inputs and outputs, at least in the simplest cases.  But what if I want 
to annotate a WSDL operation, using modelReference?  Let me take OWL-S 
initially, for purposes of illustration.  If we arrange for 
modelReference to point to an OWL-S object, it would most naturally be 
an instance of the class AtomicProcess.  The instance might be named 
something like "BuyBook" or "ReserveFlight" or "RequestPurchaseOrder". 
The atomic process instance belongs *both* to an an OWL-S ontology 
(because its an atomic process) and a domain ontology (because it has to 
do with book-buying, flight-reserving, or purchase-order-requesting) at 
the same time.

Similarly, if you decide to have your operation's modelReference point 
to a UML diagram, you won't just be pointing to some domain concept.  I 
assume you will most likely point to a StateChart diagram.  Even if we 
allow ourselves to think loosely of UML as an ontology language, it 
would be counterintuitive to say that a StateChart diagram represents 
*just* an element in a domain ontology.  A StateChart diagram represents 
a particular notion of process, just as the OWL-S AtomicProcess 
represents a particular notion of process.  A StateChart diagram will 
also have some domain-specific aspects incorporated into it.  So it's 
like the OWL-S examples (Buybook, etc) in that it captures *both* some 
aspects of service/process ontology and some aspects of domain ontology.

My point is that, in general, whatever you refer to with a 
modelReference on an operation, whether it's in OWL-S, UML, WSMO, or 
otherwise, the referent will capture *both* some aspects of 
service/process ontology and some aspects of domain ontology.

> Of course, WSDL-S could also be used
> to point to OWL-S parameters and processes, so there is no conflict.  There
> is only the question of how big this first step (in the WG) should be.
> I have no objection to OWL-S being an input document.  The question we need
> to ask is one of scope.  OWL-S, SWSF and WSMO all are based on having a
> formal ontology for modeling Web services, beyond the implicit one embodied
> in WSDL , so this will eventually be needed to achieve the full vision
> which the other group W3C proposes will define.   If the scope includes
> formally defining a service ontology, the WG will have to collectively
> figure out the best way to do this.

Personally I think W3C can provide a great deal more value to the WS 
community if it does define a core service ontology for use along with a 
set of WSDL semantic annotations.  It should be defined in RDF/OWL and 
would thus provide a natural bridge result between the Web service and 
Semantic Web activities at W3C.  The ontology should be provided not as 
"the" only semantic model, but as one canonical source of referents for 
the WSDL semantic annotations.  This initial core service ontology would 
not be as extensive as OWL-S is, but it would distill the low-hanging, 
well-understood fruit that is pretty much in common between WSMO, OWL-S, 
SWSF, Meteor-S (I think) and other work on Semantic Web Services, and 
which is naturally expressible in OWL.  (Preconditions and effects, in 
particular, would not be part of the initial core ontology because they 
don't get expressed naturally in OWL.)


> -- Joel
>>-----Original Message-----
>>From: Battle, Steven Andrew <steve.battle@hp.com>
>>Date: Wed, 16 Nov 2005 10:28:11 -0000
>>Message-ID: <DE62D3
> D0BDEF184FBB5089C7D387C37449B3B0@sdcexc04.emea.cpqcorp.net>
>>To: <public-sws-ig@w3.org>
>>The WSDL-S team openly acknowledge the strong relationship between
>>WSDL-S and OWL-S, so this should be clarified in the SAW working group
>>charter, which currently recognises only once source of input. One can
>>see the core elements of WSDL-S in the OWL-S submission. This identifies
>>OWL-S extensions for WSDL message (owl-s-parameter), binding and
>>operation (owl-s-process) definitions.
>>The relevant sections of the OWL-S submission can be found in section
>>This proposal is exactly along the lines of the WSDL working group
>>charter. I feel strongly that in the interests of clarity the sentence
>>below in the charter should be revised:
>>"A Member Submission, WSDL-S, related to this work, has been
>>acknowledged by W3C and should be used as one input to the Working
>>To something like:
>>"Member Submissions related to this work, WSDL-S and OWL-S (see
>>'Grounding OWL-S Services with WSDL and SOAP'), have been acknowledged
>>by W3C and should be used as input to the Working Group."
>>Steve Battle (Hewlett-Packard).
>>>-----Original Message-----
>>>From: public-sws-ig-request@w3.org
>>>[mailto:public-sws-ig-request@w3.org] On Behalf Of Carine Bournez
>>>Sent: 15 November 2005 14:15
>>>To: public-sws-ig@w3.org
>>>Cc: www-ws@w3.org
>>>Subject: [fwd] Draft charters for work on Semantics for WS
>>>This is a copy of an announcement sent last week to the W3C
>>>membership (on the member-ws@w3.org mailing list).
>>>All comments welcome! (for non members, on the
>>>public-sws-ig@w3.org mailing list). Thank you.
>>>>Following the announcement in [1], two charters have been drafted,
>>>>corresponding to the two points previously described.
>>>>The first one is a Semantics for Web Services
>>>Characterization Group.
>>>>   http://www.w3.org/2005/10/sws-charac-charter.html
>>>>It specifically includes 4 issues to discuss.
>>>>The second one is Semantic Annotations for WSDL Working Group.
>>>>   http://www.w3.org/2005/10/sa-ws-charter.html
>>>>Discussion on both these charters should happen on this
>>>mailing list
>>>>Thank you!
>>>----- End forwarded message -----
>>>Carine Bournez -+- W3C Sophia-Antipolis
>>Joel A. Farrell
>>STSM, IBM SWG Emerging Technology  (617) 693-5720
Received on Wednesday, 30 November 2005 02:03:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:54:15 UTC