W3C home > Mailing lists > Public > public-swd-wg@w3.org > February 2009

Re: Differences between SKOS and ISO standards : transitivity

From: Dan Brickley <danbri@danbri.org>
Date: Wed, 18 Feb 2009 21:11:07 +0100
Message-ID: <499C6B5B.7010200@danbri.org>
To: Simon Spero <ses@unc.edu>
Cc: Leonard Will <L.Will@willpowerinfo.co.uk>, Barbara Tillett <btil@loc.gov>, "Martha M. Yee" <marthamyee@sbcglobal.net>, "Allyson Carlyle (work)" <acarlyle@u.washington.edu>, Jane Greenberg <janeg@email.unc.edu>, "public-esw-thes@w3.org" <public-esw-thes@w3.org>, public-swd-wg@w3.org
(noting and setting aside for now Alistair's concern about terminology 
overlap re 'extension')

On 13/2/09 21:40, Simon Spero wrote:

> If the distinction is rejected, and the extension of SKOS concepts are
> butterflies, not documents, then it is entirely redundant in the face of
> OWL and its progeny.

This has always been my main fear with SKOS and its ancestors, eg. 
http://ilrt.org/discovery/2001/01/rdf-thes/  ... that the natural 
enthusiasm many have for "cleaning" the supposed vagueness or messyness 
from thesaurus-like structures could slip into an ill-planned 
reinvention of RDF, RDFS and OWL on top of ... RDF, RDFS and OWL. Which 
as you say would be rather redundant. We have no shortage of 
self-referential technology already.

So this kind of scoping reminder is very useful, although I will quibble 
with a detail. I think there are many people exploring SKOS who aren't 
so familiar with the facilities and conventions of the underlying RDF 
and OWL technology. The temptation to treat a SKOS "Butterfly" concept 
as a direct piece of modelling of the world is very strong. But the 
layer of indirection is there for a reason: it bridges the world with 
its various descriptions (in documents, data and elsewhere). And it 
gives a Web-identifier hook to one particular conceptualisation of 
Butterflies. When we want to say "some but not all things in the world 
are Butterflies" we have OWL already. SKOS can be used to say "some but 
not all the document-ish things in the world are _about_ _butterflies_", 
as well as "here is an id we can use for this particular notion of 
butterfly, so we can better talk about it later.".

The classes-properties-instances mindset embodied in OWL is very widely 
applicable, but can also be an awkward and stilted style when you try to 
use it for everything, and especially for topics/subjects. SKOS 
compliments it fairly nicely.

Re-reading "as a name of a subject, the term Butterflies refers not to 
actual butterflies but rather to the set of all indexed documents about 
butterflies." (Svenonius) ... I would step back from such a literal 
claim when considering what some SKOS concept might "refer to".

I'd much rather say that a SKOS "butterflies" concept is a social and 
technological artifact designed to help interconnect descriptions of 
butterflies, documents (and data) about butterflies, and people with 
interest or expertise relating to butterflies. I'm quite consciously 
avoiding saying what a "butterflies" concept in SKOS "refers to", 
because theories of reference are hard to choose between. Instead, I 
prefer to talk about why we bother building SKOS and what we hope can be 
achieved by it.

I don't believe a SKOS concept for "butterfly" (or "butterflies" or 
"Lepidoptera") is usefully described as literally referring to "the set 
of all documents about ..." those things. Nor to "the set of all users 
interested in ... those things". Nor "the set of all things that are 
butterflies". There are problems with taking any of those too literally. 
For me, the key value in the Svenonius quote is not that it tells us 
what these things refer to, but that it reminds us that we're in the 
business of connecting people with information. I can live quite happily 
without there being any story about what some SKOS concept refers 
directly to, so long as we emphasise its various named associations with 
documents and their topics; with user interests, needs and expertise, 
and also (hello, OWL...) with more formal descriptions of the world.



Received on Wednesday, 18 February 2009 20:11:52 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:31:55 UTC