W3C home > Mailing lists > Public > www-ws@w3.org > June 2003

Fwd: Re: order of the atomic process inputs

From: Mark Casey <markboulder@yahoo.com>
Date: Fri, 13 Jun 2003 15:24:10 -0700 (PDT)
Message-ID: <20030613222410.48513.qmail@web9601.mail.yahoo.com>
To: www-ws@w3.org

Titi,

Let the receiving process parse the names of the parameters to determine which is
which, e.g., when it parses ?departureLocation it will know that this is the location
parameter.

hope this helps,
Mark



Mark Casey, CTO
Miyian.com


--- titi@cs-gw.utcluj.ro wrote:
> Date: Fri, 13 Jun 2003 16:13:08 +0300 (EEST)
> From: <titi@cs-gw.utcluj.ro>
> To: <www-ws@w3.org>
> CC: <drew.mcdermott@yale.edu>
> Subject: Re: order of the atomic process inputs
> 
> 
> 
> Hi!
> 
> << Is there any mechanism in OWL-S such that when having several inputs for
> << an atomic process a certain order among them can be set?
> << The use of a RDF-Sequence (changing the look of the process ontology)
> << and sort the inputs based on some attribute (e.g input ID) would be a
> << solution?
> 
> < If you have names for the inputs, can't you sort them any way you
> < want?  Perhaps I am misunderstanding exactly what the problem is.
> 
> I try to give another view of the problem:
> 
> A service publishes its Process Ontology in which an atomic process
> checkFlight,having as inputs departureLocation and arivalLocation is defined.
> An agent,having planning capabilities finds this Process specification and
> tries to build a planning domain, by translating the atomic processes and
> their inputs to planner operators(an operator consists of an operator name
> and its parameters,e.g.(checkFlight ?departureLocation ?arrivalLocation)).
> When parsing the Process Ontology, using e.g. Jena, and translating the
> atomic process to the operator name and its inputs to operator's parameters,
> I get either
> (checkFlight ?departureLocation ?arrivalLocation)
> or
> (checkFlight ?arrivalLocation ?departureLocation)
> 
> Supplementary information is needed for the agent which translates the
> atomic process, such that it should generate an operator like (checkFlight
> ?departureLocation ?arrivalLocation) and not an operator like (checkFlight
> ?arrivalLocation ?departureLocation) as the service interprets the first
> parameter as a departure location and the second parameter as an arrival
> location. Should this kind of information be supplied in the
> Process Ontology by using a RDF-Sequence?
> This will change look of the process ontology by adding something like:
>   <rdf:Seq rdf:about="Params">
>     <rdf:_1 rdf:resource="#departureLocation"/>
>     <rdf:_2 rdf:resource="#arrivalLocation"/>
>   </rdf:Seq>
> 
> Or how should this information be provided?
> 
> Thank you,
> titi
> 
Received on Friday, 13 June 2003 18:25:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:43 GMT