[ckan-discuss] CKAN sparql endpoint

[William Waites]

* [2011-03-28 13:13:31 +0100] Rufus Pollock <rufus.pollock at okfn.org> écrit:

] @Will: can you give a bit more info esp on the SPARQL endpoint?

I'll preface this by saying that the CKAN RDF infrastructure
has been neglected a bit over the past couple of months and
it will take a small amount of work to pull it up to date, which
I'll try to do over the next week or two.

Firstly, CKAN has no dedicated SPARQL endpoint. The store
http://semantic.ckan.net/sparql is a small virtuoso install
that is running on the laboratory box. CKAN data is not the
only data in there, there is also the data.gov.uk catalogue
as well as miscellaneous other things.

Due to some infrastructure changes at data.gov.uk, that 
data is growing stale and hasn't been updated in a while.

The CKAN RDF is basicaly a DCat representation of the 
catalogue with some special cases for common extras 
fields and resource types that we know about. Particularly
if there are extras used by the LOD cloud the DCat 
representation gets augmented with voiD, but there are
other special cases too.

Richard Cyganiak and I had agreed some time ago now
to actually strip out these augmentations and just have
us generate a basic DCat description. The LOD cloud
maintainers will generate the voiD part.

Friedrich and I had discussed different ways of handling
the tags and extras in a generic way, removing the special
cases and this should be done too. The main idea is to
introduce dedicated predicates in a CKAN namespace for 
extras. This means that we really need some community
maintained definitions for these predicates, ideally using
something like Semantic Media-Wiki to collaboratively
develop the ontology. This is not terribly different from
what, for example, Open Street Map does. It would be really
great if wiki.ckan.net could be upgraded to SMW to support
this use.

If you want to work with a regularly updated dump of
the contents of semantic.ckan.net (just the CKAN part)
you can use http://semantic.ckan.net/catalogue.{n3,nt,rdf}
it isn't very big, and should be the most straightforward
way to find out what is in there.

Hope this helps,
-w
-- 
William Waites                <mailto:ww at styx.org>
http://river.styx.org/ww/        <sip:ww at styx.org>
F4B3 39BF E775 CF42 0BAB  3DF0 BE40 A6DF B06F FD45