Re: WebOnt General Requirements Subgroup - Initial E-mail

[...]

> About inconsistency... I'm happy with
> a requirement that our language should
> (a) have the ability to express
> inconsistent facts (e.g. using disjoint)
> and (b) the specification should have
> clear semantics for what's an inconsistency
> and what's not, but I don't want to
> go as far as (c) that it be straightforward, computationally,
> to determine, given any two expressions
> in our language, whether they're consistent.
>
> i.e. the sort of use cases I have in mind
> for a requirement about inconsistencies
> are, for example, catching common
> ontology-design errors (flag empty
> classes, flag empty classes that have
> things in them ;-), or well-known
> errors in the use of some ontology,
> like perhaps units mismatches in
> a space-ship-parts ontology.
>
> But I don't have any use cases that
> motivate a general decidability requirement.

my thoughts *intersect* with that
kind of inconsistency requirement
I think the examples I gave in
http://lists.w3.org/Archives/Public/www-webont-wg/2001Nov/0126.html
are *consistent* with that :-)

--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

Received on Monday, 10 December 2001 17:24:13 UTC