% notes from JC telecon 12/09/03 % by Benjamin Grosof participants: Mike Dean Benjamin Grosof Harold Boley Ian Horrocks Sandro Hawke Peter Patel-Schneider Deborah McGuinness agenda: o general news from W3C -- wrt SWRL feedback and formation of Rules WG, list discussion mode of www-rdf-rules tends to magnify any disagreements/criticisms; so far there don't appear to be major or widespread objections to SWRL punchline: we need to keep proceeding, e.g., "with all deliberate speed" o plan for NO TELECONS on 12/23 and 12/30 o next steps on SWRL - might actively solicit feedback from DAML'ers . wrt basic directions: . wrt more specifics: tools and use cases would help - tools and use cases needed . e.g., use cases / examples work on rules for services by Benjamin, Said, Massimo Paolucci . e.g., tools work by . Benjamin on SweetRules with Andi Eberhart, Said, and (past by) Boris Motik; expect new iteration of prototype in early 2004 . Jena . JTP by Deborah . Hoolet by Dimitri Sarkov, Sean Bechhofer o datatype comparisons and arithmetic Q by Mike: what's the best place to find specification of SCLP RuleML, including the EBNF syntax? A by Benj: see paper on my website: to appear in journal Electronic Commerce Research and Applications "Representing E-Commerce Rules Via Situated Courteous Logic Programs in RuleML"; also see the EBNF syntax specifications posted to JC in about May 2003 (discussion of naming issues wrt XML Schema; Peter had some points) discussion of sums, variable appearances within lists, aggregations (e.g., totals, averages), cwm limitations in that regard issue of closing off, named individuals vs. other (e.g., existential) individuals problem of indefiniteness interacting with aggregations Benj: in database and logic programs literature, aggregations are well understood to be closely related to closing off in manner of negation as failure Ian: we could push some of this into queries rather than premises; extend existing systems in that fashion Benj: yes; but often is useful to define aggregates in premises, e.g., similar to materialized views in databases, (e.g., a Morningstar 5* fund is one that is top 10 percentile over the last 5 years); can do this in manner of logic programs with NAF/aggregations