Re: Ontology "views"/"perspetcives"

> Here is the problem.  You have a (logical) ontology and
> many users that would like to customize the ontology for
> their use (have their own labels, ignore some properties,
> prefer certain properties).

You're basically talking about extensiblity and evolution. The best
way that this can be achieved in RDFS/DAML is by having a very loose
model, making sure that there are extensiblity classes. The problem
with extensions in ontologies is getting current processors to
understand the extensions enough that they can suitably understand
them. We have been trying to make sure that in EARL [1], extensions to
the current vocabulary can be at least partially recognized by current
EARL agents. This hasn't been easy, because people are not all that
comfortable with Semantic Web tools such as CWM, yet.

Some more details are scribbled down in [2]. You'll have to trawl
through the unintersting stuff, I'm afraid.

As for a standard means of going about this, the inference terms in
DAML+OIL etc. should be able to help. If you define a suitable super
class/property, then when people make sub terms, it's a simple step to
infer stuff like:-

   { ?x ?y ?z . ?y rdfs:subPropertyOf ?p } log:implies { ?x ?p ?z } .

Being able to partially understand extensions by converting them into
a form that you more-or-less understand is a simple goal of the
Semantic Web.

Cheers,

[1] http://www.w3.org/2001/03/earl/
[2] http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Dec/0038

--
Kindest Regards,
Sean B. Palmer
@prefix : <http://purl.org/net/swn#> .
:Sean :homepage <http://purl.org/net/sbp/> .

Received on Friday, 14 December 2001 12:29:29 UTC