[ckan-dev] Notes from metadata meetup / next meetup on Groups

Tue Mar 27 18:00:56 UTC 2012

Below are the notes from the recent CKAN online meetup on Metadata
Standards, taken from the etherpad: <
http://ckan.okfnpad.org/meetup-2012-03-15>.

Remember that there is a meetup on CKAN groups / Organizations scheduled
for this Thursday (5pm UK time / 6pm Central Europe / 1600 UTC). See the
announcement at <http://ckan.org/2012/03/26/groups-meetup/>. If you are
interested in attending, sign up on the etherpad: <
http://ckan.okfnpad.org/meetup-2012-03-29>.

Mark Wainwright

*
*
*Community meetup: Catalogue metadata standards, Mar 15th 2012*

There is a skeleton wiki page at http://wiki.ckan.org/Metadata_Standards (feel
free to edit it).
[Edo Plantinga (data.overheid.nl):] Also see
http://wiki.ckan.org/Data_Formats. Referrals needed?

*Participants*

Skype host: Mark Wainwright (m.wainwright) - CKAN Community Co-ordinator
James Gardner (mic problems - so mainly just listening)
Richard Cyganiak (richard.cyganiak) - Researcher at DERI - dcat, Linked
Data, etc
Adrià Mercader (adria.mercader) - CKAN development - harvesting & federation
Thomas Schandl (comfycat) - Semantic Web co - colleague of Martin
Kaltenboeck
Ludovic Coullet (lcoullet.ebiz)  - Working in a Drupal web agency with some
open data projects linked to Ckan (with Pascal Romain)
Pascal Romain (keronos)

*Apologies*

Chris Musialek (chris.musialek)

*Agenda*

Brief intro from participants (name, organization, interest).
Topics of interest - see below.

*Topic list*

Please add any topics you'd like to talk/hear/ask about!
Write your name under any topics that you are interested in.

*Experience with publicdata.eu*

   - http://publicdata.eu/ an experiment as part of the LOD2 project

   - federates catalogues from across Europe

   - ~13k datasets

   - most harvested from other CKAN instances from around Europe

   - harvesting mechanism relies on CKAN API

   - tried to harvest from other data catalogs too

   - many data catalogs don't have APIs, so often we screenscrape

   - screenscrapers fail often due to changes to the sites

   - so we are successful on the datasets that have stable APIs, but not so
   much on other catalogs

   - Example RDF output: http://publicdata.eu/rdf/package/abandoned-vehicles

*Controlled vocabularies in CKAN*

   - LOD2 project could add controlled vocabularies to CKAN

   - This is actually already work in progress (for EC data portal)

   - Possibly available in next release

*Inspire*

   - directive from EU on publishing geospatial data

   - Inspire has a “catalog service” based on the CSW standard and ISO
   19193 standard

*Status report on DCAT*

   - DCAT is a proposed interoperability standard for data catalogues

   - A metadata model (doesn't define the API at present)

   - RDF based

   - Started with DERI, moved to W3C

   - http://www.w3.org/2011/gld/wiki/Data_Catalog_Vocabulary

   - http://dvcs.w3.org/hg/gld/raw-file/default/dcat/index.html

   - http://www.w3.org/TR/dcat/ (future permanent URL)

   - Close to getting a permanent public status

   - Not guaranteed stable, but won't change too much as things already
   built on it

   - ADMS: http://joinup.ec.europa.eu/asset/adms/description (an extension
   of DCAT)

   - DCAT-based harvesting mechanism from the Edinburgh workshop:
   http://wiki.okfn.org/OpenDataCatalogues/2#Proposed_Harvest_Mechanism

   - If people implement it & feedback to the WG it would help it along

Questions:

   - Is the controlled vocabularies management going to evolve in DCAT? Or
   do we just need to include skos elements in our implementation of it?

   - It just uses SKOS, so that should be fine. This won't change.

   - If I understood it right the thesaurus is defined at the catalogue
   level and the terms of indexation at the dataset level ?

   - Thesaurus (= skos:ConceptScheme) is defined at catalog level, but can
   be re-used from somewhere else (existing thesauri)

   - themeTaxonomy is used to define the skosConceptScheme ?

   - It's the attribute of the catalog that points to the skos:ConceptScheme
    ok

   - dcat:theme is used to define the the Skos:concept used?

   - It's the attribute of the dataset that points to the skos:Concept ok

   - So we can use several controlled voc in the same dcat file (example
   eurovoc and Inspire)

   - Yes!

*Push notifications*

Any specs for these, eg realtime changes being synced between instances?

   - XMPP (a.k.a. Jabber)?

   - James: Would be good to link that up with some of the CKAN API in the
   future.

*Use cases for metadata standards*

   - W3C draft document:
   http://dvcs.w3.org/hg/gld/raw-file/default/dcat-ucr/index.html

   - Send comments to the working group!

*C**KAN** connector for Drupal 7*

   - This topic was added by Ludovic, who had to leave early, so it was not
   discussed. Feel free to send questions to the ckan-dev list (
   ckan-dev at lists.okfn.org).

-- 
Mark Wainwright, CKAN Community Co-ordinator
Open Knowledge Foundation http://okfn.org/
Skype: m.wainwright
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/ckan-dev/attachments/20120327/957d189e/attachment.html>