RE: validating Hydra documents (and ontologies)

On Sunday, August 31, 2014 11:11 PM, Dimitris Kontokostas
> Hi Markus, all,
> 
> The main idea behind RDFUnit is that vocabularies, ontologies,
> knowledge bases and applications should be associated by test cases
> that assure a basic level of quality.
> 
> For the first 2, we provide test generators that ensure the data are
> valid in a CWA OWL setting and we convert automatically owl / rdfs
> axioms to test cases (not full owl/rdfs support yet) We also have
> generators for DSP and Resource Shapes constraints checking (not 100%
> compliance either yet)

Great work!


> In addition to the automatic test case generation we support manual
> test cases attached to vocabularies & ontologies that are loaded when
> ever we test against a schema

So if we would add a rut:testSuite link to our test suite to the Hydra vocabulary definition, RDFUnit would automatically be able to go and fetch the test suite, right?


> e.g. we define in geo that the lat / lon should be in the range of
> [-90,90] & [-180,180] and when ever someone uses -s geo as an option
> we load both the automatic and the manual test cases in the validation
> process [1] We did some initial work on manual test cases in a few
> vocabs like dbpedia, geo, lgdo, skos, lemon & nif [2] that are
> available for everyone that uses RDFUnit and loaded by default. In the
> same manner you could define logical constraints for the hydra vocab
> that would also be available by default when someone tests a hydra
> dataset.

Cool


> I also updated the documentation a bit on the github wiki (there is a
> lot more to add) but this should give you the basics
> https://github.com/AKSW/RDFUnit/wiki
> 
> In addition, what we do atm, is to enable targeted validators that are
> pre-configured with specific schemas and make domain specific
> validation easier https://github.com/AKSW/RDFUnit/wiki/LIB

Yeah, that would be very valuable. Not just for the Hydra Core Vocabulary but also for Linked Data Fragments... I'm sure Ruben already talked to you about that though :-)


> The old demo is back online but I am working on a new version that I
> will announce soon. Let me know if you have any questions.

Great. Please CC public-hydra when you do so!


Thanks,
Markus


> 
> Best, Dimitris
> 
> 
> [1] https://github.com/AKSW/RDFUnit/blob/master/data/tests/Manual/www.w3.org/2003/01/geo/wgs84_pos/geo.tests.Manual.ttl
> [2] https://github.com/AKSW/RDFUnit/blob/master/data/tests/Manual/
> 
> 
> 
> On Thu, Aug 7, 2014 at 4:44 PM, Dimitris Kontokostas <kontokostas@informatik.uni-leipzig.de> wrote:
> 
> On Aug 7, 2014 5:22 PM, "Markus Lanthaler" <markus.lanthaler@gmx.net> wrote:
> >
> > On Thursday, August 07, 2014 4:15 PM, Dimitris Kontokostas wrote:
> > > On Aug 7, 2014 2:12 PM, "Luca Matteis" <lmatteis@gmail.com> wrote:
> > > >
> > > > On Thu, Aug 7, 2014 at 12:27 PM, Markus Lanthaler
> > > > <markus.lanthaler@gmx.net> wrote:
> > > > > Very cool. Is there a way to test this or at least see the rules
> > > > > that would
> > > > > be run without installing anything?
> > > >
> > > > This is why I always opt for JS-browser-based apps. You can use them
> > > > without having to install anything and you don't have to fear servers
> > > > going down (as they usually do).
> > >
> > > I know but time is the only constraint here:-)
> > > I'll put it back online in a few days. Either way, the web demo is
> > > lacking many features that are available on CLI and works only with
> > > sparql endpoints. A new more light weight  version is in progress but
> > > will probably take a while to release.
> >
> > OK
> >
> >
> > > Any feedback is more than welcome
> >
> I think it would be very helpful if you could you briefly explain what you tools is actually doing (e.g., are there fixed validation rules or, if not, > based on what criteria are they generated?) and how we could help to make your tool more useful for users of Hydra.
> Right now I'm in a tiny keyboard and I can only point you to the publications at the bottom of our homepage :-) 
> I'll come back with a detailed response in ~10 days.
> Best,
> Dimitris
> >
> >
> > Thanks,
> > Markus
> >
> >
> > --
> > Markus Lanthaler
> > @markuslanthaler
> >
> >
> >
> >
> 
> 
> 
> 
> -- 
> Dimitris Kontokostas
> Department of Computer Science, University of Leipzig
> Research Group: http://aksw.org
> Homepage:http://aksw.org/DimitrisKontokostas

Received on Thursday, 4 September 2014 11:15:41 UTC