- From: Shadi Abou-Zahra <shadi@w3.org>
- Date: Mon, 11 Aug 2014 08:39:53 +0200
- To: ERT WG <public-wai-ert@w3.org>
- CC: Eric Prud'hommeaux <eric@w3.org>
Hi folks, Ref: http://w3.org/brief/Mzgy I was playing around with "RDF Shape Expressions" [1], which seems to (finally) provide the capability to write logical rules (constraints) that we previously tried to do with OWL and RIF (without success). - [1] http://www.w3.org/2001/sw/wiki/ShEx Using EricP's ShEx Demo I was able to build a very rudimentary EARL validator that shows a possibility of using this new technology. Note that the Demo and all RDF processing is fully client-based as well. Of course, the basic set of rules that I created have several caveats: #1. It does not (yet) validate nodes of the supplementary EARL modules HTTP-in-RDF, Content-in-RDF, and Pointers-in-RDF; #2. It assumes description fields are XML Literals though they could also be XSD Strings - this actually doesn't seem to be very clearly defined in DC, FOAF, and DOAP, which is a bit of an issue; #3. Languages for strings and literals are not being considered which can lead to false positives in the validator; #4. The validator only reports matches rather than non-matches, which is a consequence of using this particular Demo. I encourage you to have a look at RDF Shape Expressions. I think that with a refined set of rules and some test cases we could propose solid "exit criteria" for a W3C Candidate Recommendation stage for EARL 1.0. Best, Shadi -- Shadi Abou-Zahra - http://www.w3.org/People/shadi/ Activity Lead, W3C/WAI International Program Office Evaluation and Repair Tools Working Group (ERT WG) Research and Development Working Group (RDWG)
Received on Monday, 11 August 2014 06:40:23 UTC