W3C home > Mailing lists > Public > public-lld@w3.org > December 2010

Re: SemWeb terminology page

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Fri, 3 Dec 2010 09:51:40 +0100
Message-ID: <4CF8AF9C.2020105@few.vu.nl>
To: public-lld <public-lld@w3.org>
Hi everyone,

Perhaps a bit of explanation on what we discussed in Cologne :  http://lists.w3.org/Archives/Public/public-xg-lld/2010Dec/0005.html

For both groups we felt that "vocabulary" was not a good match. In the semantic web terminology we have "RDF vocabularies" which are the sets of classes and properties used to create data [1]. This would clash with using "vocabulary" for the other group: a "controlled vocabulary" represented in RDF (using SKOS for example) is not an RDF vocabulary.

We explored using qualifiers to mitigate that, but while "value" was perhaps the best fit with "vocabulary" for the first group, we felt it still had important problems. One may expect to find string values (literal) in "value vocabularies", while what is expected in the LLD environment would stuff that in RDF would be represented as "fully-fledged" resources.

Further, in the LLD environment a lot of things may appear in group 1: thesauri, classifications, but also authority lists for persons, places, events, or even objects. Also, stuff like the entire dbPedia set, or a list of foaf:Persons. It may be far-stretched to use "vocabulary" for the latter. Which is why we came with the quite flexible "dataset", preferably combined with a qualifier like "reference" or "organized" (we did not have any preference though--both?).

In fact, as was observed, belonging to group 1 is truly a matter of *function*, how that stuff is being used. It is the fact that someone or some organization says: "here's a set of resources, I've curated it and believe it is fit for re-use". And conversely, the fact that others start re-use it in their data.

Note that there is no strict disjointness between the two groups, as Marcia wrote it. And this matches well the flexibility of "dataset". An the classes and properties in an RDF ontology (that is, an "RDF vocabulary" ;-) ) may be used as a KOS for creating descriptions, if some application needs it.



[1] http://www.w3.org/TR/rdf-schema/ says: "This specification describes how to use RDF to describe RDF vocabularies."

> I have to say that I think that 'value vocabulary' and 'element vocabulary' make the most sense for me.  As soon as we start doing things like 'library vocabulary' and 'semantic web vocabulary' we start defining what we're talking about in terms of different communities of practice rather than function, and it all starts looking like the US Congress (not a good thing, for those of you who don't have to watch it up close).
> Diane
> On 12/2/10 9:37 PM, Thomas Baker wrote:
>> Hi Jodi,
>> On Thu, Dec 02, 2010 at 10:13:47PM +0000, Jodi Schneider wrote:
>>> The use of 'vocabulary' with different modifiers seems
>>> doomed to fail. That's because, for me, I find it difficult
>>> to mentally distinguish 'Value vocabulary' and 'element
>>> vocabulary'. The idea of a 'library vocabulary' and 'semantic
>>> web vocabulary' is just barely understandable enough for me
>>> to handle.
>> I'm not sure I follow. "Semantic Web vocabulary" I can
>> understand, but is the use of "vocabulary" uniform enough
>> in the library world for it to make sense to speak of a
>> "library vocabulary"?
>>> I'd be very, very happy if someone could propose an
>>> alternative which didn't use 'vocabulary' twice. I fear
>>> abbreviation as well as the assumption that, oh, yeah, we know
>>> what vocabularies are (with different resultant assumptions
>>> depending on one's background).
>> "Metadata element set" has been proposed as an equivalent
>> to "element vocabulary". I would argue, however, that
>> speaking of "element vocabularies" and "value vocabularies"
>> usefully underlines the fact that in the Linked Data context,
>> the two types are comparable as Semantic Web (or RDF)
>> "vocabularies".
>> I'm convinced that there are no terms we could come up with
>> that would not evoke the wrong associations for _someone_.
>> This is an exercise in coming up with terms that roughly evoke
>> the right sorts of things for as many people as possible.
>> Whatever terms we choose, we then have to define them clearly
>> and concisely, up-front -- as in "when we say 'vocabulary',
>> we mean...".
>> But I hear your strong view on this and would be interested
>> to know if others share the fear that using "vocabulary" in
>> this broader sense would simply prove to be too confusing to
>> too many people.
>> Tom
Received on Friday, 3 December 2010 08:51:12 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:27:42 UTC