Calling a spade a spade, and a key a key

Sandro Hawke just posted a nice exposition of an RDF problem
that contained the following text:

> >The first is a description of a web page and the second is a
> >description of the W3C.   They are each unambiguously identified, but
> >different reverse-functional properties are used.

The phrase "reverse-function" property refers to what are more
commonly know as "keys".  The OWL spec uses the term
"inverseFunctional" in place of the term (non-compound) "key" to
describe an RDF property that serves as a key for instances
belonging to the domain of that property.  Within the small community
of description logic developers, the equivalence of "inverseFunctional"
and "key" are well understood.  Outside of that community, relatively
few individuals will figure out this mapping until someone schools
them in the meaning of the arcane OWL terminology.

The fact that OWL omits the notion of a compound key from its
spec is a major oversight.  The fact that they omit the term "key" from the
spec is also unfortunate.  This kind of disregard for common modelling
practices will retard the acceptance of RDF by a larger community.  We
can partially overcome the obstacles imposed by OWL by adopting
conventional terminology in our own correspondence, briefings, etc.

Cheers, Bob

Received on Friday, 13 December 2002 13:51:04 UTC