RE: [ACTION 11] Initial DCAT to meta-share mapping

Thanx for the explanations! Indeed the foaf:primaryTopic sounds bizarre!

Just as a general comment, the current ms vocabulary does not include all the elements of the original ms xml schema; for instance, services have not been mapped and the distribution was also lost in the mapping from xml to rdf (for various reasons, good enough). Anyway, I am now trying to check the overall schema also and how it will fit to the final rdf.

So, for the dataset property, can we add it only to the subclasses of ms:LanguageResource rather than the ms:LanguageResource itself, at least until a final decision is made on whether we'll also include tools/services as a subclass thereof?

Best,

Penny

 

From: johnmccrae@gmail.com [mailto:johnmccrae@gmail.com] On Behalf Of John P. McCrae
Sent: Wednesday, September 10, 2014 12:57 PM
To: Penny Labropoulou
Cc: Dave Lewis; public-ld4lt@w3.org
Subject: Re: [ACTION 11] Initial DCAT to meta-share mapping

 

Hi Penny,

 

The mappings does not replace any existing properties, simply adding new ones to make the datasets queriable using the DCAT vocabulary, hence adding Dataset to all the language resources. The metashare vocabulary still seems to be a bit messy with Services, but for the data I currently have no services are declared so I am not so worried about the mappings.

 

foaf:primaryTopic is the property used by DCAT to link a catalog record to a dataset, which also seems bizarre to me but is the standard.

 

Regards,

John

 

On Wed, Sep 10, 2014 at 11:37 AM, Penny Labropoulou <penny@ilsp.gr <mailto:penny@ilsp.gr> > wrote:

Hi John and all!

 

Maybe the confusion comes from me, sorry for that: I thought the two documents were complimentary, with the wiki providing the principles for the mapping and the spreadsheet including the detailed mapping. So, I decided to put in the wiki the rationale and (since I though the mapping was under discussion) check the detailed spreadsheet once the general decision was made; I had already added some comments inside the spreadsheet. Anyway, do I have time to send in some comments for the final mapping by tomorrow?

BTW, I 'm not sure what the mapping of the various ms.BabelNet, ms.ComputationalLexicon etc. to dcat.Dataset is: are they going to be subclasses of dcat:dataset or replaced by it or ???? There's an ongoing discussion as regards the mapping of ms:LanguageResource to dcat:Dataset as there's an overlap between them.

 

And I was having a better look at the licensing module to make sure all the MS elements are there and correctly mapped to odrl and other vocabularies; is this ok?

 

Best,

Penny

 

BTW, I had a quick look at the github and I think there are some "copy-and-paste" errors – e.g. in the mapping of MS.metadataCreationDate I see FOAF.primaryTopic as well as the dct.issued.

 

 

From: johnmccrae@gmail.com <mailto:johnmccrae@gmail.com>  [mailto:johnmccrae@gmail.com <mailto:johnmccrae@gmail.com> ] On Behalf Of John P. McCrae
Sent: Wednesday, September 10, 2014 11:55 AM
To: Penny Labropoulou
Cc: Dave Lewis; public-ld4lt@w3.org <mailto:public-ld4lt@w3.org> 
Subject: Re: [ACTION 11] Initial DCAT to meta-share mapping

 

Hi,

There seems to be some duplication here as the metashare to DCAT mapping to the spreadsheet here:


https://docs.google.com/spreadsheets/d/15SE4_qAqYFostmD52uKxpkCPZh1f5TrPeoXKNTlDYpQ/edit#gid=0

Can one of the LD4LT/LIDER chairs please decide which document we are using??

 

The mapping in the spreadsheet is already implemented here:


https://github.com/liderproject/metadata-harvesting/blob/master/metashare/fix_metashare.py

 

Regards,
John

 

On Wed, Sep 3, 2014 at 2:53 PM, Penny Labropoulou <penny@ilsp.gr <mailto:penny@ilsp.gr> > wrote:

Hi Dave and all.
I've made some comments (in italics to easily spot) at the wiki for the
MetaShare to DCAT mapping. Please, note the following:
- the proposal looks only at dcat:dataset and dcat:distribution but there
are also dcat:catalog and dcat:catalogRecord. MetaShare has not looked at
catalogs at all, but some of the elements can be mapped to
dcat:catalogRecord properties. Should we look into this also?
- Some of the comments refer to the XML implementation of MetaShare (i.e.
elements that have not been included in the current version of MS/RDF-OWL)
but I have added them as they will be of use for the final XML to RDF
conversion of the original MetaShare schema.
Best,
Penny



-----Original Message-----
From: Dave Lewis [mailto:dave.lewis@cs.tcd.ie <mailto:dave.lewis@cs.tcd.ie> ]
Sent: Thursday, July 31, 2014 3:44 PM
To: public-ld4lt@w3.org <mailto:public-ld4lt@w3.org> 
Subject: [ACTION 11] Initial DCAT to meta-share mapping

Hi all,
I've made an initial analysis of how we can make use DCAT in revising the
meta-share vocabulary at:
https://www.w3.org/community/ld4lt/wiki/DCAT_MetaShare_Mapping

Some points we can discuss on the call shortly:

1) classifying ms:LanguageResource as a dcat:Dataset - see:
https://www.w3.org/community/ld4lt/wiki/DCAT_MetaShare_Mapping#Make_Language
Resource_a_DCAT_Dataset

2) separating meta-data about the resource from meta-data about its
accessible forms or distributions - see:
https://www.w3.org/community/ld4lt/wiki/DCAT_MetaShare_Mapping#Separate_Lang
uageResource_metadata_from_metadata_of_its_Accessible_Form

3) (which I haven't documented yet) is using dcat:theme attribute for
LanguageResource properties that comply to a specific scheme or taxonomy
using skos:concept and skos:ConceptScheme. This would make sense for
ms:domain, but also perhaps for making explicit the subclass taxonomy
LanguageResource, e.g. Corpus etc.

Talk to you shortly,
Dave