owl semantics

Hi,

I've a query arrising from the owl semantics page [1]. Section 2 defines a 
vocabulary (signature), in terms of owl stuff (classes, properties, 
datatypes, ...). It also defines interpretations of ontologies. I was unsure 
about a couple of points.

1) there is no formal definition of the relationship between an ontology and a 
vocabulary. One possibility is that each ontology can be transformed into a 
vocabulary that mentions exactly those things used in the ontology. Another 
is that there is one vocabulary that contains *everything* and that any 
ontology only uses some of. Of course, a range of possibilities exist between 
these two extremes. The former seems more likely to me.

2) the datatype domain N_D is defined as containing all of the base xsd string 
sub-types - but does this hold if a particular ontology does not refer to one 
of these? This would tend to support the seccond possible interpretation 
above, namely that there is one vocabulary of everything.

Perhaps the wording is confusing me, and instead these xsd types must be 
supported as datatypes by any complient system? That is, using these 
datatypes in a vocabulary is guaranteed to in itself not be grounds for 
classifying the vocabulary as outside the range supported by owl 1.1?



Matthew

1 http://www.w3.org/Submission/owl11-semantics/

Received on Monday, 23 July 2007 13:47:17 UTC