RE: [RIF] homework for 10/17 telecon (R2ML...)

On Mon, 2006-10-16 at 22:47 +0200, Gerd Wagner wrote:
> > > Please have a look at the XML Schema of our REWERSE Rule Markup
> > > Language (R2ML) at
> > > http://oxygen.informatik.tu-cottbus.de/rewerse-i1/?q=node/6
> >
> > Interesting... lots of stuff there... I think the
> > "EBNF abstract syntax" is the main thing I want to study;
> > I'm having trouble finding it. Help?
> 
> We've defined the abstract syntax of R2ML with the help
> of a UML/MOF model (as proposed by the OMG):
> http://oxygen.informatik.tu-cottbus.de/R2ML/0.4/metamodel/R2MLv0.4.htm
> 
> MOF is more abstract (and in some sense more expressive) than
> ordinary EBNF, and it can be easily mapped to EBNF.

I'm somewhat familiar with UML and MOF; the W3C TAG is using
UML a bit, and I have written some code to convert the source
of the diagrams to OWL (i.e. RDF/XML using OWL terms).

See: Using RDF and OWL to model language evolution
2006-02-15
http://dig.csail.mit.edu/breadcrumbs/node/87


What is the source of that metamodel/R2MLv0.4.htm stuff?
I see HTML files and images; is there a MOF source file
somewhere that I could study?

> But for those who prefer to read EBNF (and I'm afraid
> this is the majority with a W3C group), we'll make an
> EBNF for the typed condition language of R2ML by tomorrow.

I look forward to it.
> 
> > The first example I see starts with an incorrect namespace
> > declaration...
> >
> > <r2ml:ProductionRule
> > xmlns:ex="org.drools.examples.TroubleTicketExample"
> >  -- http://oxygen.informatik.tu-cottbus.de/rewerse-i1/?q=node/17
> >
> > The value of a namespace attribute is a URI; "org.drools..." is not.
> 
> Yes, sorry, we didn't map the Java package name to a corresponding
> URI - will be fixed by tomorrow :-)
> 
> > Do you have any example of using URIs from RDF/OWL documents
> > in R2ML rules?
> 
> R2ML rules can be based on an external vocabulary in the
> form of an RDFS vocabulary or OWL ontology. See the MOF 
> diagram below that defines the possibility of zero or more external
> vocabularies.

OK... I guess that makes sense... I'll have to study it some more.

I'm more interested in practical examples...

> We have also a running example where an external vocabulary
> in the form of the OWL ontology 
> http://oxygen.informatik.tu-cottbus.de/TestVoc.owl
> is combined with R2ML rules that are being verbalized, see
> http://oxygen.informatik.tu-cottbus.de/verbalization/index.jsp

Yes, that is quite interesting... (especially since I'm involved
in a research project involving proofs/explanations, where
this verbalizing stuff is quite relevant. see
http://dig.csail.mit.edu/TAMI/ )

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Monday, 16 October 2006 21:49:56 UTC