Re: Quick "last call" before CR for Data Shapes' SHACL

Hello

I would like to share this SHACL use-case with the group :

The European Legislation Identifier initiative (ELI) (
http://eur-lex.europa.eu/eli-register/about.html) is an initiative that
encourages European national legal publishers to share reusable metadata
about legislation; the goals are to provide an easier access to law, a
common legislation metadata language (the ELI ontology at
http://publications.europa.eu/mdr/eli/index.html), and to interlink
national and european legislations. All of this is done with URIs and RDFa
metadata included in legal publishing websites. It is implemented by
national legal portals in France, Luxembourg, Italy, Danemark, Norway,
Finland, Ireland, and in the European legislation portal
http://eur-lex.europa.eu; implementation status details can be read at
http://eur-lex.europa.eu/eli-register/implementation.html.

To improve the quality of shared legislation metadata, the ELI taskforce
has been experimenting with SHACL to build an "ELI validator", that enables
participant countries to assess the semantic validity of their RDFa markup
against a set of shapes derived from the ELI ontology. The validator is
currently deployed in its prototype version at
http://labs.sparna.fr/eli-validator/. It relies on TopQuadrant's
implementation of SHACL (https://github.com/TopQuadrant/shacl). A few
specificities of this validator are :

   - definition of SHACL shapes from an Excel spreadsheet (see
   http://labs.sparna.fr/eli-validator/registry), to allow quick
   tool-independant editing of shapes; the excel spreadsheet is then converted
   to RDF using an online conversion service (
   http://labs.sparna.fr/skos-play/convert);
   - validation or RDFa metadata after parsing of the webpage;
   - ability to display a set of SHACL shapes in a webpage (for example
   http://labs.sparna.fr/eli-validator/display?shapes=http://datashapes.org/schema
   );
   - dispay the validation report broken down by shapes/constraints (and
   not by validation results); this requires, as an additional step after
   validation, to check whether each shape/constraint actually matched some
   target in the validated data, in order to distinguish constraints that
   "passed" (the constraint matched some target and no validation result was
   produced from the constraint) from constraints that "didn't matched" (the
   constraint didn't matched any target in the validated data);

Disclaimer : the ELI validator is based on the version of SHACL from
January 2017. It is currently in a testing phase.

Best Regards

Thomas



2017-03-03 20:54 GMT+01:00 Sandro Hawke <sandro@w3.org>:

> [re-sending with hopefully correct list name this time.   First try, with
> cross-posting, is at
> https://lists.w3.org/Archives/Public/semantic-web/2017Mar/0014.html]
>
> The RDF Data Shapes Working Group is nearing the end of its charter and
> wants to do one last quick check to see if there's anything that really
> MUST be fixed before going to Candidate Recommendation.
>
> New draft published today: https://www.w3.org/TR/shacl/
>
> As the draft says:
>
>    - Please provide your *feedback by Friday, 17 March 2017* at the
>    latest. If you need more time than that, please let us know what you need.
>    - If you have made earlier comments that you don't consider as having
>    been addressed, please tell us again.
>    - We are especially interested in plans to implement. Comments simply
>    indicating plans to use this work are also helpful.
>    - You can either submit your feedback as an issue on our GitHub
>    repository <https://github.com/w3c/data-shapes/issues> (preferred) or
>    via the public mailing list
>    <https://lists.w3.org/Archives/Public/public-rdf-shapes/>.
>
> Thanks!
>
>       -- Sandro Hawke, W3C, on behalf of the RDF Data Shapes Working Group
>
>
>



-- 

*Thomas Francart* -* SPARNA*
Web de *données* | Architecture de l'*information* | Accès aux
*connaissances*
blog : blog.sparna.fr, site : sparna.fr, linkedin :
fr.linkedin.com/in/thomasfrancart
tel :  +33 (0)6.71.11.25.97, skype : francartthomas