Joint Committee Minutes 21 August 2001

This page summarizes the telecon for the Joint US/EU Committee on Agent Markup Languages held from 1300 to 1400 PDT on 21 August 2001. These minutes were prepared by the chairman, and were approved during the 28 August telecon.



No announcements this week.

Last Week's Minutes

The minutes from August 14 were approved by those present.

Rules Proposal

Benjamin introduced the proposal that he and Ian wrote up following the meeting at IJCAI. This is a starting point; more details (e.g. lefthand-only vs. righthand-only restrictions) should emerge over the next few weeks and should establish the theoretical basis. It will probably take a step beyond that to get a concrete language proposal (syntax, etc.). Benjamin and Ian are adopting an approach of "interative deepening" of the proposal. In particular, the KR-fusion approach requires more working out.

Harold Boley suggested in the IRC channel using an extension of CARIN-ALLNR.

Ian noted that we need to investigate more to determine if the combined language is in fact less than FOL. We don't want to spend lots of effort reinventing FOL.

Pat Hayes asked for a definition of "rule" as used in this proposal. ACTION (Benjamin): provide a definition of "rule"

Horn axioms (conjunctive premise in body, conclusion in head) are the point of departure. Some other constructs are expressively reducible.

More rule examples would be desirable. Richard offered to identify Ontolingua ontologies that include axioms.

Query and Results/Justifications Language

Richard opened the discussion on query languages. He doesn't want to wait for a rule language. He mentioned Ian's IEEE paper on Query Languages for Description Logics.

Stanford KSL has a strawman query proposal in-house. The premise is a DAML+OIL knowledge base. The query consists of DAML+OIL statements + variables. ACTION (Richard): distribute this proposal to

Pat opined that while we'd like to find a natural way to fit query and rules together, convergence is desirable, but non-convergence wouldn't be fatal.

Lynn noted that non-convergence could be a problem, e.g. if the query language was more powerful than the rule language. Ian gave the example that you can't currently express uncle (based on brother and child) in DAML+OIL.

Pat suggested that query processing should involve entailment.

There was some discussion of handling of unnamed individuals. Deb addressed this problem in her thesis work with Borgida. She also referred people to a paper Asking Queries about Frames.

There was some discussion of a staged approach. Ben suggested that we start with subsumption and existence and build from there. It may make sense to release an Interim Query Language for querying current DAML+OIL knowledge bases, and extend this to support rules later.

We should be at least "mutually aware" of related efforts including RQL, SquishQL, and XQL. There was a RDF Query BOF at SWWS that included 5 proposals.

DAML+OIL (March 2001) Copyright

We discussed the copyright issues associated with submitting DAML+OIL to W3C (presumably as a W3C Note). Everyone wants to make this happen, but we don't know how to proceed. We also want to ensure that derivative works will be unconstrained.

W3C basically wants to ensure that the material will be freely usable by it and other organizations. It doesn't require transferring the copyright to W3C (like a journal publication). Folks/organizations joining WebOnt will have to relinquish or declare any existing Intellectual Property rights, so they're presumably already covered.

Ian suggested just making DAML+OIL public domain or releasing it under terms such as the GNU Public License (GPL).

ACTION (Mike): reply to the W3C folks and ask what else we need to do

Next Steps

We continued the discussion on evolution of the Joint Committee begun last week.

Ian and others noted that we need a useful interaction between the rule and query languages, and that this is a good place to do it.

Please let Mike know if you'd like to take this opportunity to gracefully resign from the Joint Committee.

It was previously suggested that we add a member of the DAML-S Coalition to the Joint Committee. Rather than picking someone ourselves, we'll ask the DAML-S folks to suggest a liaison with their group. ACTION (Mike): contact the DAML-S folks.

We'll plan to have a group vote next week on inviting Benjamin Grosof, Richard Fikes, Harold Boley, and a DAML-S representative (if named) to officially join the Joint Committee. Please send any negative votes or other concerns to Mike beforehand.

Next Week

The following items are currently on the agenda for next week:


Kelly Barber's raw notes
last week's minutes
Joint Committee home page

$Id: 2001-08-21.html,v 1.10 2001/09/03 21:35:50 mdean Exp $