W3C home > Mailing lists > Public > public-rif-wg@w3.org > January 2006

[RIF] Re: Work on JRules abstract grammar

From: Christian de Sainte Marie <csma@ilog.fr>
Date: Tue, 17 Jan 2006 19:32:22 +0100
Message-ID: <43CD3836.8080405@ilog.fr>
To: public-rif-wg@w3.org

All,

During the conf'call today, I asked for comments and opinion on how to 
best follow-up on Philippe's suggestion - that (it could be benefitial 
that) participants in this WG share meta-models/abstract syntaxes of 
their rule systems - but then we reached the time limit, and I had an 
action set on me to move that thread to the emal.

I had the time to see Ed, Said, Harold on the queue before we split and 
I had the impression that there where more...

So, let the discussion start...

(btw the way, I do not know what the subject tag for that thread should 
be? Is it RAF or do we need a new subject code?)

Christian

Philippe Bonnard wrote:

> Hello,
> 
> I would like to mention briefly an exercise that I did at ILOG with the
> aim of trying to describe the ILOG Rules Language (IRL) used by ILOG
> JRules. We think that such an exercise may be of interest for the rest
> of the RIF Working Group.  The exercise consists simply in using
> XML/RDF/RDFS/OWL/PRR to represent as faithfully as possible a subset of IRL
> syntax.
> 
> We found that this work has helped us get a better idea of the important
> features of our language that we expect to be covered by RIF for PhaseI,
> and the way we could see them expressed in a RIF.
> 
> It would perhaps be beneficial if others in the WG did (or have done)
> the same sort of exercise with their own system or language. Then,
> comparing/contrasting our homeworks might give us a concrete base to
> lead us to an actual RIF.
> 
> If the rest of the WG is interested, I would be glad to share what I did
> with the group. See the attached document for some more details.
> 
> Best regards,
> 
> Philippe.
> 
> 
> ------------------------------------------------------------------------
> 
> In order to have an appreciation of how our JRules production rule
> language could be expressed using a RIF, I have extracted from the
> language a partial abstract grammar consisting of some intermediate
> nodes that I deemed relevant for the task.  
> 
> I limited the extraction to
> the expected features of the RIF Phase I (i.e., essentially syntax). In
> so doing, I deliberately chose to remain at some level of abstraction to
> allow for a wider family of languages. For example, the "Statement" and
> "Action" nodes haved not been developed.  
> 
> For the other nodes, I limited
> the elaboration to the essential (value expression) since I thought it
> was too early to be lost into the details of the grammar.  I presented
> the result of this exercise internally at ILOG: an abstract (sub-)
> grammar as a UML diagram specifying the structure of the rules. Using
> the "nodes" of this abstract grammar, I could write a basic ontology -
> one using RDF(S) alone and one using OWL.  
> 
> Finally, I also proposed a
> tentative XML concrete syntax for such a RIF-like format in the form of
> an XML schema in order to study the extensibility capability of the
> abstract language.
> 
> 
> ------------------------------------------------------------------------
> 
> 
> ------------------------------------------------------------------------
> 
Received on Tuesday, 17 January 2006 18:32:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:33:26 GMT