Re: Developing a standard for an ontology registry/repository
Leo, Paola, Jack, et al.,
LO> As John mentioned early on in his message, there is an Open
> Ontology Repository effort at Ontolog, which is an open community
> of practice. You might join and post to the ontolog OOR list.
Yes, it is important not to split the efforts into two competing
projects. I agree that the basic work should be done in the OOR
group and that the SUO can be used to develop an IEEE standard
(if and when a standard seems appropriate).
PDM> but 'ontology requirements for sharing and reusing them' are
> one thing, the 'project requirement for a repository/registry'
> is another
Yes, but a registry/repository has always been considered
essential for any system of sharing and reusing. The two
have been discussed together for ages upon ages.
JR> I agree with your statements except for the precise meaning
> of 'requirements.' I sense that you use the term to mean more
> like 'post conditions for satisfaction' whereas most people
> (and majority of PM's) use the term to mean the functions
> and features of an envisioned system regardless of the symptoms
> of the problematic situation.
I used the term in the vague general sense of including everything
that people usually include in the word 'requirements'. The exact
rules for what should be included or excluded can be determined
and revised as people start putting things in the list.
In any case, this discussion should be moved to the OOR list,
as Leo suggested. My primary concern, as I have stated all along,
is that a fundamental requirement is that the ontologies in the
repository should have explicit generalization/specialization
links. That makes it into a hierarchy (partial ordering), which
is a subset of the lattice of theories.