FW: SKOS Reference editors' update

Hi all,

Sean & I had a phone call this morning, to discuss progress on the SKOS
Reference. Here's a quick summary of our discussion.

== Label Relations ==

Al's been thinking about Antoine's comments on label relations and the
SKOS-XL sketch. There are three alternative patterns for representing label
relations, being (1) n-ary relations between plain literals (currently in
the SKOS Reference); (2) n-ary relations between some other class of
resources (e.g. xl:Label); and (3) direct (binary) relations between some
other class of resources. We don't seem to be in a position to recommend any
one in favour of the others. So maybe we should move support for all three
patterns outside the SKOS Reference to an extension? Al will try to write a
message on this.

== SKOS & Imports ==

Sean had a chat with Simon Jupp recently, regarding implementation of SKOS
within an OWL environment. Simon noted that the use of owl:imports between
SKOS concept schemes pushes you into OWL Full. We're not sure what to do
here, given that OWL imports is the only existing mechanism for importing
RDF graphs, and given the likely resistance against "rolling our own"
imports mechanism. Sean will try to send a message to the list on this
topic.

== Formal Schema ==

Sean has a first draft of a formal schema for SKOS, written in Turtle. We
had a quick discussion, and agreed first priority is to ensure the axioms
agree with the prose definitions in the SKOS Reference. Al noted that use of
SKOS documentation properties will make the schema OWL Full, maybe possible
to have a SKOS formal schema as an OWL DL ontology? 

Regarding the namespace, Sean has for now assumed we'll be using a new
namespace. We had a quick discussion about the namespace issue, Al has no
strong favourite from the two options, Sean (mildly?) favours a new
namespace.

== Mapping Relations ==

Al plans to have a look at the question of mapping relations. We briefly
discussed the four scenarios, and when to use semantic relations versus
mapping relations, the scenarios being (1) describing internal structure of
a concept scheme; (2) mapping between different concept schemes; (3) linking
two or more concept schemes to make one "big" concept scheme; (4) adding
structure to someone else's concept scheme. Al will try to write a message
on this. 

Regards,

Alistair.

--
Alistair Miles
Senior Computing Officer
Image Bioinformatics Research Group
Department of Zoology
The Tinbergen Building
University of Oxford
South Parks Road
Oxford
OX1 3PS
United Kingdom
Web: http://purl.org/net/aliman
Email: alistair.miles@zoo.ox.ac.uk
Tel: +44 (0)1865 281993

Received on Monday, 14 April 2008 16:02:09 UTC