W3C home > Mailing lists > Public > public-esw-thes@w3.org > June 2008

RE : Meaning of "broaderTransitive"

From: Antoine Isaac <Antoine.Isaac@KB.nl>
Date: Sun, 8 Jun 2008 15:31:12 +0200
Message-ID: <68C22185DB90CA41A5ACBD8E834C5ECD04953DCC@goofy.wpakb.kb.nl>
To: "Leonard Will" <L.Will@willpowerinfo.co.uk>, <public-esw-thes@w3.org>

Dear Leonard,

> It seems to me that confusion about "broaderTransitive" and
> "narrowerTransitive", especially their being wider in meaning than
> "broader" and "narrower", arises from their being given inappropriate
> names.
> Calling a relationship "broaderTransitive" appears to say something
> which restricts the nature of the relationship, whereas it seems to be
> being used in SKOS just to say something about the concepts which can be
> related by it.

Creating a transitive closure of a property might appear as a restriction, but it is not if you really look at the formal semantics of (OWL property) transitivity
You can have a look at the threads
where I tried to explain this.

But I think anyway you already got it right, if I understand properly your following question:

> Am I right in thinking that on the analogy of direct relationships being
> expressed as "parent / child", the idea that "broaderTransitive /
> narrowerTransitive" tries to express is "ancestor / descendant" ?


> If this is the case, then it is clear that "broaderParent" is a specific
> case of "broaderAncestor", and these names would be less confusing. If
> you don't like them, how about "broaderIndirect" for "broaderAncestor",
> leaving "broader" on its own to express "broaderParent", for which it
> has generally been used in thesauri?
> I still find "A broader B" to be ambiguous, and would like to see it
> expressed as "hasBroaderConcept" to avoid confusion with
> "isBroaderConceptOf". These would appear quite neatly as
> "hasBroaderParent" and "hasBroaderAncestor" or even "hasParent" and
> "hasAncestor".
> Similarly for narrower . . .

I cannot blame you for thinking this, I quite agree for myself. The problem is that skos:broader has already quite widely used.
But the 'ancestor' idea might be quite relevant, even though a bit pleonastic
(but maybe in the end redundancy is what we really need: broader-in-the-sense-of-ancestor-and-by-the-way-it-should-also-be-used-for-classification-scheme-specialization ;-)


Received on Sunday, 8 June 2008 13:32:24 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:45:48 UTC