Re: blog: semantic dissonance in uniprot

On Mon, 30 Mar 2009 09:11:48 -0700, Michel_Dumontier  
<Michel_Dumontier@carleton.ca> wrote:


> (although it's unclear from
> your website whether services are *actually* described in an OWL
> document or they just refer to OWL types).

Both/either  (assuming that I understood your question)

The service interfaces are defined as OWL classes.  If you chose to point  
to an existing OWL class (with or without a formal definition of it's  
property restrictions) that's fine, though the while system works better  
if you don't do that.


> People still have to commit to the corresponding semantics at some
> point. Right?


People have to commit to an agreement on the intent/meaning of  
predicates.  Nothing more.

M