% notes from JC telecon 10/07/03
% by Benjamin Grosof

o Rules draft:  

warning label:  
- people reasonably happy with where we are on it after Benj's comments
- Benj to do more drafting based on his edits
- more edits on the example:  
  . give more context, a full rule rather than an atom
  . maybe give an abtract syntax version instead or in addition

people reasonably happy with Annotation being added cf. OWL
- Benj then all: later we can add rule-specific meta-info such as rule-labels, 
set up namespaces, etc.
- Deborah then all:  let's mention that those kinds of things are coming,
to give headsup, get suggestions 
- Ian:  will want to name OWL axioms too, 
including for proofs/justifications etc. 
- Benj:  instance statements too
- Peter:  it's clumsy in RDF:  to name an OWL axiom or rule requires
naming a collection of statements, thus have to quote and unquote a 
lot of stuff since; it's easier with rules since the RDF is treated just
as syntax
- Mike, yes, have a top-level object "Rule".  
- cf. the document draft:  
don't expect the Rules entailments to be a superset of 
RDF entailments
- Benj:  yes, this is something we've discussed several times before


o DAML PI Mtg Rules session 

plan:  
- intro by Mike Dean
- Ian and Peter can present their draft
- Harold will present syntax
- Said will talk about applications
- Benj will lead the discussion, help prep the intro with Mike incl.
directions for next steps (see bit more about that below)

also want to inject some technical discussion into the joint 
session with OntoWeb
- short presentations on both Rules and SWSI
. for Rules:  OWL-Rules and RuleML concrete syntax ; 
. about 30min in the agenda there; let's keep the presentation to 15min 
. mention requirements and use cases
. related rules work:  Jena2, SCL, OMG
. key decisions etc.

we can request follow-up discussion on www-rdf-rules

wrt names:  "OWL Rules" may be kept reserved as a name for future W3C


o prioritization for next steps:

- wrt timeout for objections:  
Fri. 10/10/03 6p PDT
Mike will send out to DAML etc. lists, and direct discussion to 
www-rdf-rules

- directions:  
built-ins
NAF
prioritization 
etc.
(cf. our previous discussions)


o OMG:  led by Said (Gerd was there too)

OMG is considering what we're doing for their model-driven architecture,
including RuleML and OWL ontologies  

emphasis there on model-driven architecture, for enterprise level,
including processes and rules and now ontologies.
they now have a SIG on ontologies.
they are looking at a lot of issues in mapping OWL or even RDFS,
and for rules and rule engines in general.
the topic Said and Gerd were involved in was introducing rules and RuleML
to OMG people in general.  
semantics of business rules is one group.
another group deals with issues of production rules.
typical companies involved:  ILOG and Blaze and a number of researchers
including from Northface University (Utah).
Gerd:  that's a new university that specializes in software engineering,
funded by a number of companies like IBM and Microsoft.  They are 
interested in using business rules for software engineering, thus
quite interested in rules.  
Said and Gerd did a presentation on rules and RuleML including relationship
to MOF model.  MOF (Meta Object Facility) model is a meta-language for
specifying abstract syntax (did I get this right?), related to subset of UML.
There's continued discussion with the SIGs at OMG.
There are two RFP's out
- modeling business rules
- integrating production rules technology with UML

Mike:  are the two RFP's synchronized with each other?
Said:  an interesting point
Mike:  what's the timeline?
Gerd:  submission deadline is I think in February

action item for Said and Gerd:  please send pointers to the RFP's and 
about MOF, maybe write a paragraph

Computer Associates has a production rule product called Ion

Said: RuleML can help with the need -- recognized there -- for
interoperability between different production rule systems

Benj:  yes, there's already work on this both by Said and by Benj+others
on SweetJess which translates a large fragment of SCLP RuleML back and forth
to Jess, probably the currently most important production rule system

(discussion about link at OMG to the Ontologies issues, including
overlaps in membership with DAML and OntoWeb and OWL committee)

Said:  also interesting is E-Wallet work by Norman Sadeh et al 
on combining ontologies with rules
Mike:  they have a nice paper at ISWC as well

Deb:  Jeff Bradshaw and colleagues are using for encoding policies via rules
- there are ~ two pertinent papers at ISWC