W3C home > Mailing lists > Public > semantic-web@w3.org > March 2009

Re: Vocabularies for content keyword classification?

From: Peter Krantz <peter.krantz@gmail.com>
Date: Mon, 9 Mar 2009 16:50:10 +0100
Message-ID: <7b9ad66d0903090850g66c53647g23aa2968466126ce@mail.gmail.com>
To: Semantic Web <semantic-web@w3.org>
Thank you all for your suggestions! I received many suggestions both
on and off-list. Here are our plans based on your feedback:

> 1. What vocabularies have you used to define a single free text
> keyword? We'd rather use an existing term instead of inventing our
> own.

For "raw" keywords that exist in the current unstructured dataset
we'll make publishers include the data in an RDF metadata file by
using one or more http://purl.org/dc/elements/1.1/subject elements,
one per keyword:

<dces:subject xml:lang="en">Occupational health</dces:subject>
<dces:subject xml:lang="en">Some other term</dces:subject>
...

> 2. For those organizations that have made an effort to structure their
> list of keywords, what would you advise them use to make it easy for
> them to publish it online and make references instead of free text
> keywords in their document metadata?

This seems to be an area where everyone agrees that SKOS is the way to
go. One issue though is that the current spec isn't finished yet and
in http://www.w3.org/TR/2008/WD-skos-reference-20080829/#L1302 it says
that the namespace URI may change back to
<http://www.w3.org/2004/02/skos/core#>. There seems to be a number of
early adopters out there. What SKOS namespace URI do you recommend?

Regards,

Peter Krantz
Stockholm, Sweden
Received on Monday, 9 March 2009 15:52:52 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 21:45:28 GMT