RE: [tm-pubsubj] Contradicting a PSI

["Bernard Vatant" <bernard.vatant@mondeca.com>]

Patrick, Peter

I see your points. Maybe "is inconsistent with" is a better choice than
"contradicts"
to capture that. At least it speaks to people with some background in logic
...

Something like :

"The assertions made by an user application about a subject identified by a
PSI
should be consistent with the assertions made about this subject in the
subject indicator.

In the case assertions made by the subject indicator are expressed in some
formal language,
consistency can be checked by logical tools using this language.

In any case, the formal or unformal description of the subject should
provide as much precision as possible
about assertions constituting the subject definition (such as type and
other constitutive properties),
to avoid PSI users to be misled in inconsistent uses."


Bernard Vatant
Senior Consultant
Knowledge Engineering
Mondeca - www.mondeca.com
bernard.vatant@mondeca.com


> -----Message d'origine-----
> De : Peter Flynn [mailto:peter@silmaril.ie]
> Envoyé : lundi 8 mars 2004 23:33
> À : tm-pubsubj@lists.oasis-open.org
> Objet : RE: [tm-pubsubj] Contradicting a PSI
>
>
> On Mon, 2004-03-08 at 20:54, Bernard Vatant wrote:
> > Patrick
> >
> > I meant (I think Lars Marius also) a contradiction in the
> following sense
> >
> > Publisher declares a PSI to identify a subject A
> > User uses it as a PSI to identify a topic X, of which actual
> subject is B,
> > not identical to A.
> >
> > 'A is not identical to B' means 'it exists some assertion true
> for B that
> > is false for A'.
> > This is the simplest way to say two subjects are different :
> you can say
> > something which is true for one and false for the other.
> >
> > This is in which the user 'contradicts' the publisher definition.
>
> I think we need to be very careful before adopting this term.
> Contradiction (in English) means 180&degree; opposition, not
> merely a small variation or *some* assertion true for B which is
> false for A: it means *all* assertions which were true for B are
> false for A.
>
> Perhaps "colliding" or "conflicting" PSIs would be closer?
>
> ///Peter
>
>
>
> To unsubscribe from this mailing list (and be removed from the
> roster of the OASIS TC), go to
> http://www.oasis-open.org/apps/org/workgroup/tm-pubsubj/members/l
eave_workgroup.php.