Re: UDDI and semantics: CMU OWL-S/UDDI Mapping

Ugo,

Ugo Corda wrote:

> Did you find this front-end approach sufficient? Or do you think that 
> additions to the existing UDDI data structures intended to directly 
> support RDF/OWL information would bring substantial benefits?
>
Unfortunately, I do not have a good answer, and I will have to think 
about your question some more.  We took UDDI as a done deal, and we 
tried to fit DAML-S into it, instead of redoing UDDI.  Two places where 
some semantic annotation may help are the tModels and the category bag,  
which seem to be the main ways to retrieve information from the 
repository.   For other objects it all depends on what questions UDDI 
users may want to ask.  For example consider the binding,  if the goal 
is just to save binding information and retrieve it then using OWL may 
not add anything, but if the goal is to be able to ask which web 
services use a binding compatible with HTTP constrained by some security 
parameters,  then WLthere may be a case for O

Received on Tuesday, 2 December 2003 14:17:50 UTC