my notes from today's 4/29/03 JC telecon discussion on RDF/OWL as possible syntax for Rules

From: Benjamin Grosof ([email protected])
Date: 04/29/03

  • Next message: Mike Dean: "eXtensible Rule Markup Language (XRML)"
    Hi folks,
    here are some notes I took today.  Mike, you may find them useful to merge 
    into your minutes, too.
    % notes from JC telecon 4/29/03:  aspects on Rules, incl. OWL/RDF as syntax
    % by Benjamin Grosof
    1. Benjamin presented quick ovw of his msg earlier this wk
    on Approaches to Negation in RuleML,
    to help kickstart email discussion
    (discussion of some other OWL stuff)
    2. continuing discussion on possible OWL syntax for RuleML:
    Mike D:  seems there's a parallel possibly to
    why there's as yet only an XML (rather than RDF) syntax for DQL:
    "glass container" principle that don't want container to influence
    the contents
    Sandro:  actually, my DQL implementation is in terms of SOAP
    Sandro:  analogy:  are we arguing about whether OWL syntax for
    RuleML would it be an A or D, vs. would it be a D vs. F;
    personally feel it probably would be C vs. D
    Benjamin:  some have argued that rules should be on top of RDF,
    e.g., W3C Plenary and Team, RDF devotees
    what are arguments for why have rules on top of RDF syntax?:
    (at least triples, and then possibly OWL too):
    - RDF is supposed to be cleaner as a basis for KR, generally
    - unifying the tool set, e.g./also for:
    . querying, generally
    . unifying with RDF Query
    . meta-querying, e.g., find all rules that conclude a particular predicate
    - "work well and play well with others", where others are part of
    SW set of standards
    - wrt OWL:  have ontology(ies) of rulebases, rules, atoms, variables,
    hierarchy of rule sublanguages based on expressive and syntactic features
    - Mike D:  could embed rules in the same page as ontologies
    Sandro:  maybe there just needs to be a well-defined mapping to RDF,
    rather than actually be in RDF syntax
    Pat:  on top of semantics of RDF vs. on top of just syntax of RDF
    Stefan and Benjamin:  basically, just syntactically
    Pat:  as a notation for graphs, RDF is not great, esp. its XML encoding
    discussion of consensus of meaning going beyond what the formalism
    can directly capture, e.g., meaning of primitive terms used,
    social context of purposes
    Sandro:  e.g., comments by Timbl at end of DAML PI Mtg in Portland 10-2002
    about meaning of a URI
    some discussion of truth predicate in RDF, meta-languages/contexts
    and meaning
    Sandro:  real RDF engines will be doing business logic, rules, ontologies,
    using URI's
    Peter:  then they are not RDF engines
    Sandro:  maybe say they are RDF-speaking
    Peter:  no, they're something else;
    related to Social Meaning Interest Group that's supposed to be starting in May
    Prof. Benjamin Grosof
    Web Technologies for E-Commerce, Business Policies, E-Contracting, Rules, 
    XML, Agents, Semantic Web Services
    MIT Sloan School of Management, Information Technology group or

    This archive was generated by hypermail 2.1.4 : 04/29/03 EST