- From: Jakob Voss <jakob.voss@gbv.de>
- Date: Mon, 11 Sep 2006 15:28:40 +0200
- To: public-esw-thes@w3.org
Hi,
Andrew Houghton from OCLC gave me some good input about the encoding of
DDC with SKOS, so I updated the
Quick Guide to Publishing a Classification Scheme on the Semantic Web
http://esw.w3.org/topic/SkosDev/ClassificationPubGuide
To a second tentative draft, 11 September 2006.
These are the changes:
== Notations ==
ISO 693-2's language code "art" for articifical languages is
inappropriate for notations. Fortunately there is a new language code
"zxx" for non-linguistic. Notations in SKOS should always use this
language code. Note that if you specify no language code then there is
no way to determine if a prefLabel is a notation or not the the language
code should be mandatory for notations.
== Non-symmetric relations ==
Using rdfs:seeAlso for non-symmetric relations is probably
inappropriate. dcterms:references (an element refinement of dc:relation)
should be used instead.
== Extended Notes ==
For textual class description XHTML is probably the best choice. Note
that XHTMl can also include RDF with RDFa. I'm not sure if this example
is valid:
<skos:Concept rdf:about="http://www.ams.org/msc/76-xx.html">
<skos:prefLabel xml:lang="zxx">76-xx</skos:prefLabel>
<skos:note>
<rdf:value rdf:parseType="Literal">
<p xmlns="http://www.w3.org/1999/xhtml"
xmlns:dcterms="http://purl.org/dc/terms/">
For general continuum mechanics, see
<a href="http://www.ams.org/msc/74Axx.html"
rel="dcterms:references">74Axx</a>,
or other parts of
<a href="http://www.ams.org/msc/74-xx.html"
rel="dcterms:references">74-xx</a>
</p>
</rdf:value>
</skos:note>
</skos:Concept>
Maybe we can encode more of SKOS with RDFa so we can get rid of the
cruel RDF/XML encoding that is broken by design anyway.
== Subject counts ==
SKOS allows to specify subject indexing of single records with
skos:subject, skos:isSubjectOf, skos:primarySubject, and
skos:isPrimarySubjectOf. However for many applications the total number
of records indexed with a specific concept is needed while the actual
relations may be unknown. I propose a new property
skos:skos:hasSubjectCount and a class skos:skos:SubjectCount - please
have a look at the wiki page for examples.
Greetings,
Jakob
P.S: Anyone else coming to the NKOS workshop at ECDL next week?
Received on Monday, 11 September 2006 13:27:34 UTC