Re: rdfs:seeAlso Re: Semantic E-mail

An idea about how to "define" a vocabulary - and cope with versioning - that
I came up with in discussion with Alistair Miles yesterday (who was good
enough to write it up at on the
esw Wiki)

Roughly, the idea is to declare an OWL Ontology for each term we use, then
declare the OWL onotology we want to actually describe. This allows us to use
OWL with the single-term ontologies to describe when a new term obsoletes an
old one - since an Ontology that replaces another one with something
different probably isn't ever entirely compatible.

A little more inline...

On Sat, 10 Jul 2004, Laurian Gridinoc wrote:

>On Sat, 10 Jul 2004 09:41:03 +0100, Richard Newman
><> wrote:
>> It tells a crawler "get this file for more information about this
>> resource". So, for instance data that seems to make sense.
>RDF Schema:
>"rdfs:seeAlso is an instance of rdf:Property that is used to indicate
>a resource that might provide additional information about the subject
>resource", which applied to my construct:
>But how can I interpret the recommendation doesn't count; all it
>counts is the consensus on how a vocabulary can be constructed
>referring similar properties in other vocabularies rather than using
>them directly (mixing vocabularies) in order to promote partial

Hmm. I don't think that it would be a bad thing to use rdfs:seeAlso. But I
think people have interpreted pretty broadly the idea of "the resource they
are providing more information about".

>I would like to mix vocabularies, but I won't know how to choose them
>-- I mean I would choose elements which semantics satisfies my needs,
>but how can I know that I'm not then `speaking' a dead language (ex. a
>forgotten draft, etc)?

Well, if you are speaking it then it isn't dead :-) If it meets your needs, I
think it makes sense to use it (unless there is evidence that the information
about it is likely to actually disappear (see about why this is a bad thing...) and
there is nothing to stop you describing the resource yourself in any way you
want (although doing anything that might change the meaning is probably a bad

>There is no popularity engine for schemas, and hoping that in the
>future there will be servers that will provide the vocabulary
>equivalence information is like hoping for link servers to happen.

Well, there are such systems being developed. As far as I know there isn't
anyone who has put the kind of resources a Web search engine uses at the
disposal of an RDF crawler, but people are building things that could do this
(SemWalker from W3C is one approach, others provide straightforward access to
their collected information...)



Received on Saturday, 10 July 2004 08:42:59 UTC