Re: SemWeb terminology page

On Tue, Nov 2, 2010 at 3:33 PM, Haffner, Alexander <A.Haffner@d-nb.de> wrote:
> Hi Mark,
>
> Good point!
>
>> About terminology:
>>
>> I understand that FOAF, FRAD, ... are called "vocabularies" by this
>> group (although I would term them "schemas" or "metadata schemas").

So RDF had a technology naively called RDF Schema in 1998/9. At the
same time the XML world decided they wanted a 'schema language' to
replace/augment DTDs. This made the naming of RDFS a bit awkward, and
we (RDF WGs then) were gently encouraged to think of a name that
didn't look like it overlapped in purpose with this [much more
industry-approved] XML stuff. People got quite upset about such things
at the time (RDFS was nearly a W3C REC in 1999, then the XML Schema WG
found out and got upset (since if RDFS was a schema 'standard' then
maybe they'd be forced to use RDF somehow; and so RDFS REC was
postponed while things were sorted out! probably needed it anyway...).
See http://www.w3.org/TR/schema-arch and nearby.

This is how RDFS sort of became "RDF Vocabulary Description Language"
in later drafts, http://www.w3.org/TR/rdf-schema/

At the same time around ~2000, people who called themselves
'ontologists', and called these things ontologies, noticed RDFS and
that with a few additions, it could be a language for ontologies on
the Web. So RDFS was a very basic ontology language. Renaming it 'RDF
basic ontology language' would've been seen as undermining the then
new initiatives to produce a richer RDF-based ontology language. So we
stuck with RDFS.

> I like also the RDA, FRBRer etc. terminology "element sets"

Dublin Core used to talk about having an 'element set'.

Then XML took over (same story :) and in XML, element meant something
very particular. And since DC favoured RDF for it's expression in XML,
and since RDF/XML allows metadata 'elements' to be expressed as XML
markup *attributes* (eg. dc:title="...." not <dc:title>), DC moved
away from talk of elements, towards the term 'term'.

>> But what are AAT, LCSH, ... called, if not "vocabularies" (at least I
>> hope a separate term is reserved for terminological resources).

I quite like 'vocabulary' as it covers schema, ontology, metadata set,
and also SKOS/thesaurus stuff too.

Generally the whole situation is a mess, and can only be understood
sociologically / historically. Asking whether eg. 'schemas' and
'ontologies' are the same or different doesn't get us very far. Asking
about which communities used which terms maybe gets us a bit
further...

cheers,

Dan

Received on Tuesday, 2 November 2010 14:52:39 UTC