W3C home > Mailing lists > Public > public-esw-thes@w3.org > October 2005

Re: [PORT] new editor's working draft of SKOS Core Vocab Spec

From: Sue Ellen Wright <sellenwright@gmail.com>
Date: Sun, 9 Oct 2005 13:14:05 -0400
Message-ID: <e35499310510091014p3aaf60ffv195a5574e3e4d32a@mail.gmail.com>
To: Dan Brickley <danbri@w3.org>, Alan Melby <melbyak@yahoo.com>, Klaus-Dirk2 Schmitz <klaus.schmitz@fh-koeln.de>
Cc: tiago.murakami@itau.com.br, "Miles, AJ (Alistair)" <A.J.Miles@rl.ac.uk>, public-esw-thes@w3.org
Dear all,
 I think Tiago has touched on a very serious issue that bothers us in the
terminology field as well. We see controlled vocabularies as a very specific
kind of resource that are used for the purpose of information organization
and retrieval. Their purpose is not necessarily at all to provide linguistic
resources for people creating texts, for instance, which is what most
terminological resources do. So there is a significant difference between a
controlled vocabulary and even a standardized terminology. My guide on what
a controlled vocabulary is the new NISO Z39.19 draft standard for Controlled
Vocabularies, which includes thesauri, semantic rings, and authority files
(subject headings). (During the balloting process, this document was readily
available on the web, but I didn't just find it right now. If anyone wants
it, I can find a place to put it, with the caveat that it is undoubtedly
changing now during the post-ballot editing phase.)
 The "term" *terminology* shows up in the context of controlled
vocabularies, but here it is used polysemically to refer to thesauri and
thesaurus-like knowledge organization schemes. Terminologies (of which a
folksonomy could indeed be a specific kind of manifestation) have quite a
different purpose. They are designed to document real usage in discourse,
usually specialized discourse of some sort. They can serve as merely
descriptive information or as recommendations or even prescriptive
information. Of course, controlled vocabularies can be used in this way, and
terminologies can be used for information retrieval, but these applications
are variations on the original intention of the resources in question.
 Another factor to take into consideration is that in many cases very
different communities of practice are involved in creating the resources and
hence the resources exhibit different metadata structures. Folksonomies tend
to be lexicographically oriented in many cases, although they can be
terminological. Whereas formal special language terminologies do indeed
include concept-system apparatus, many folksonomies and dialect dictionaries
do not, in part because of their lexicographical orientation. Elements of
the folksonomy tradition can even cross-over into specialized terminologies,
such as in the multi-layered register-related synonyms stored in
public-service oriented multilingual medical resources.
 The critical criterion remains that controlled vocabularies are created for
information organization and retrieval. They are used to document objects
stored in systems, whether those systems are physical like libraries and
museums, or purely electronic. Lexicographies and terminologies, on the
other hand, document real language and its usage in the
*uncontrolled*environment of both general and specialized discourse.
They may well provide
information on conceptual relations, but the purpose behind this exercise is
to guide users in selecting appropriate words and terms to use in real
discourse and to impose rigorous control of definition structures and
inter-conceptual relations. Information retrieval from collections of real
or digital objects is only a secondary (and sometimes cumbersome)
application of these kinds of terminologies.
 Bye for now
Sue Ellen

 On 10/9/05, Dan Brickley <danbri@w3.org> wrote:
> tiago.murakami@itau.com.br wrote:
> >Hi All,
> >
> >There is a problem: Folksonomies are not a Controlled Vocabulary.
> >
> >
> >
> My view is that they are controlled, just in a different way. On my blog
> I control my keywords / categories, and arrange them in a basic
> hierarchy. On flickr, I do the same with my "tags" that I assign to
> photos. In both contexts I do this with some thought for how they relate
> to the categories used by my friends and colleagues. And in both cases,
> there are tools to expose these categories in RDF/SKOS. They're
> certainly not controlled in the classic library sense, but they are
> organised; sometimes carefully, sometimes carelessly. The weblog case is
> more clearly "controlled vocabulary" than Flickr (based only on current
> UI). This is because in my blog, when I post an article via Wordpress,
> it offers me a list of my existing categories as the options for
> categorising a post. On Flickr there is a free-text entry field instead.
> But UIs can change easily: the practice in both systems leads people to
> use the same category/keyword over again.
> Short version: folksonomies are "locally-controlled vocabularies",
> perhaps?
> Dan

Sue Ellen Wright
Institute for Applied Linguistics
Kent State University
Kent OH 44242 USA
Received on Sunday, 9 October 2005 17:14:14 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 2 March 2016 13:32:06 UTC