00:01:18 Slide: Semantic Community 00:02:23 Slide: 'Definition' 00:02:55 ISO 1087-1 and 24707 (Common Logic) 00:03:01 Slide: Business Vocabulary 00:03:09 separate meaning from notation 00:03:46 non-normative SBVR structured English 00:04:06 enforcement requires rules about rules 00:04:14 Slide: Semantic Formulation 00:04:45 Slide: Seamntic Formulation of a Simple Rule 00:05:01 strongly linguistics oriented 00:05:08 integrated with formal logic 00:05:48 Grosof: RIF must be able to support communication of non-automatable rules? comment field? support? 00:06:09 e.g. must wear hard had on construction site 00:06:26 enforcement can't be automated 00:06:54 Grosof: perhaps different term that non-automatable 00:07:25 can analyze for ambiguity, inconsistency, etc. even if machine can't enforce 00:07:56 ?: rules should be in human readable form - XML is basically binary 00:08:05 who spoke? 00:08:22 Igor Mozetic 00:09:51 Speaker: Gary , Oracle 00:09:58 DB glossary 00:10:40 Slide: query 00:10:53 optimize using algebraic rewrite rules 00:11:16 Gary Hallmark, Oracle 00:11:29 Slide: view 00:12:24 Slide: integrity constraints 00:12:44 conditions that must hold at transaction boundaries 00:12:49 perhaps add glossary term for transaction 00:13:00 glossary+ transaction 00:13:27 constraints resemble queries - same syntax, different interpretation 00:13:35 Slide: triggers 00:14:24 Slide: how to evolve the KB 00:14:46 Slide: role based access control 00:14:47 security 00:14:56 in scope for RIF? 00:15:31 Grosof: use of rules for security policy? 00:15:49 Grosof: kitchen sink, without distinguishing layers 00:16:16 Grosof: is this satisfactory? 00:16:51 Gary: arbitrary security requirements, minimum result set size, etc. 00:17:19 Barkmeyer: hoping for more about transactions 00:17:53 Barkmeyer: can't evaluate rule safely until everything available - unitary change 00:18:20 Gary: rules in database vs. rules in middle tier 00:18:32 Gary: ECA rules can't impact other rules 00:18:45 There is no such verb as "to inference"! :-) 00:18:55 Speaker: Chris Welty leading open discussion 00:19:17 Welty: What are we doing? How are we doing it? 00:19:50 Boley: scalability to scope of the web? 00:20:10 Boley: hard to control or debug, particularly for active rules (production and reactive) 00:20:23 Welty: scalability is more of a requirement 00:20:44 RIF focus on interchange, not defining a new language 00:21:01 q+ 00:21:06 understanding will emerge over time 00:21:16 should be less fuzzy as a result of this meeting 00:21:28 tomorrow reserved almost exclusively for discussion of use cases and requirements 00:21:45 Boley: are we concerned about efficient execution of interchanged rules? 00:21:56 Jon Pellant 00:22:11 holger has joined #rif 00:22:32 big difference in scope for interchange of rules vs. rule instances running in live environment 00:22:47 Mala: like to see support for evolution of large rule bases 00:22:55 Welty: sounds like requirement 00:23:40 Cory Casanave: what isn't a rule? process modeling as rules about behavior? every UML element? 00:24:17 Dave Springgay: usability, human readability, authoring - outside RIF scope 00:25:05 Hassan: different conceptions of formal semantics logicians as model theory semantics - doesn't have to be model-theoretic; other formalisms possible 00:26:10 Hassan: rules discussed today mostly involve inference, vs. just flexible way to express computation (production rules) 00:26:20 Hassan: just changes of state 00:27:02 Hassan: business rules don't require inference, just agile computing - useful to distinguish evaluation/execution from forward/backward inference 00:27:44 ?: not a response but elaboration - what are we going to do in this working group 00:27:57 ?^Pascal Hitzler 00:28:11 thanx 00:30:04 Grosof: think we should identify KR expressiveness and type of generic computational tasks associated with clusters of systems and use cases, then identify superset of subsets that enables significant translation and interoperability between pairs with well-understood semantics 00:30:28 Grosof: model theory is good at specifying these functions 00:31:08 Grosof: could support multiple KRs - e.g. SWRL sits on cusp between Horn FOL and Horn LP 00:31:42 Grosof: behaviors/ground conclusions often identical 00:32:31 Grosof: a lot known theoretically - should have 1 or 2 umbrella KRs, at least for this phase 00:32:46 Grosof: perhaps annotated comments 00:33:21 Welty: clusters of existing rule languages and systems (with possible overlap) 00:33:31 Welty: classifying existing systems 00:33:53 Welty: may not achieve 100% interoperability - standard approach is to cluster things with common assumptions 00:34:38 Sandro: keeps the market for extensions simple (package per cluster) 00:34:59 -PedramAbrari 00:35:18 Is the speaker Joshua Engel? 00:35:27 yes 00:35:54 Engel: also accommodates non-rule systems - queries are a special case of rules 00:36:31 Barkmeyer: different tasks: inference for the purpose of creating knowledge vs. immediate information vs. validation of information set 00:37:03 Barkmeyer: tools tend to cross between tasks (DBMS as kitchen sink) 00:37:17 Barkmeyer: 3 related but different axes 00:37:36 Welty: not hearing any dissent on classifying systems 00:37:45 Bijan: not sure - still assimilating 00:38:40 Bijan: worried about scope - job of working group to survey existing systems? 00:39:06 Welty: necessary for interchange? 00:39:40 Sandro: sides with Bijan - those communities are here, don't have to worry about others until Last Call 00:39:55 Bijan: would rather start with rule systems represented here 00:40:47 Christian: agree - but if some community is missing, may need to look outside for extensibility 00:41:09 Sandro: how does this relate to use cases? top down vs. bottom up? 00:41:28 Sandro: meet at features (systems) or requirements (use cases) 00:42:09 Ginsberg: clustering good - ontology about rules or reasoning may be helpful 00:42:57 Ginsberg: capturing features (e.g. types of nonmon) may help avoid confusion 00:43:31 Kifer: series of languages? interchange implies multiplicity 00:43:52 Kifer: define language(s) underlying format 00:44:02 Barkmeyer: define abstract concepts 00:44:06 Kifer: and concrete syntax 00:44:16 q+ 00:44:19 q- 00:44:45 Welty: classification could drive requirements for syntax and semantics of interchange format 00:44:51 Kifer: OK 00:45:12 Bijan: use cases for rules vs. use cases for exchange 00:45:19 MDean: +1 00:45:42 Welty: most use cases for rule language not interchange between rule languages 00:46:29 Welty: use cases fairly standard in standards efforts - to assess coverage, etc. 00:46:43 Welty: no use cases focused on interchange 00:46:47 http://lists.w3.org/Archives/Public/public-rif-wg/2005Dec/0055.html 00:46:59 Dieter: use cases provide guidance on relevant features 00:47:08 Possible use cases from a interchange perspective 00:47:17 I don't agree with that :) 00:47:50 Grosof: agree, but how valuable/likely is exchange of certain rules? 00:48:10 Grosof: e.g. context required to run aprocs 00:48:45 Grosof: counter-argument is combination of local context with exchanged information - need to semantically integrate 00:49:12 Dieter: too limited notion of exchange - could be within processes/vendors locally 00:49:15 Grosof: agree 00:49:39 Paul: need to define "exchange" 00:50:22 Bijan: need to prioritize features? 00:50:57 Ginsberg: use case showing exchange discussed tomorrow 00:51:19 Ginsberg: also compare consistency of rules 00:51:24 Hassan: 1+ 00:51:50 q? 00:52:21 Welty: enumeration problem - use case for each rule engine 00:53:02 Welty: e.g. what kind of interopability can be achieved between systems with different semantics 00:53:58 q+ 00:54:03 Boley: agree - should combine use cases - compare translated results - multiple groups working on same domain (e.g. bioinformatics) 00:55:03 Hassan: translate between formalisms preserving structure (homomorphism) - PRR started with this (meta-model) 00:55:48 Hassan: same issue here - PRR approach makes continuing sense 00:56:26 homomorphism can be forgetful - exchange only patterns 00:57:08 Hassan: first describe objects in common 00:57:44 Cory: relative stupidity of meta-models can be an advantage - simplify synthesize common meta-model 00:58:21 Cory: use cases for rules have probably been accommodated in existing languages 00:59:25 David Springgay: OMG goal to identify core then extend 00:59:37 jrp has joined #rif 00:59:42 David Springgay: solves goal of interoperability between vendors 01:00:27 Dave Reynolds: what do rule vendors expect from RIF that PRR with XML serialization doesn't provide? 01:00:30 Bijan: or RuleML? 01:01:19 Paul: excellent question - PRR aimed at OMG and modelling - clear need for more real-time non-modelling interchange closely related to modelling interchange 01:02:15 Paul: RIF should be run-time interchange format based on PRR - production rule support required, otherwise irrelevant 01:02:44 Paul: RIF in XMI possible, but other formats probably better 01:03:08 Reynolds: concrete syntax for PRR would meet requirement? semantics requires? 01:03:40 q+ 01:03:46 Paul: no concrete syntax for PRR core (main meta-model) - what other concrete syntax will be required for RIF 01:04:01 q- 01:04:13 Welty: need concrete syntax for rule interchange - phase 1 for core subset 01:05:03 Paul: meta-model describes only semantics - e.g. doesn't define expression language (e.g. RDF) 01:06:03 Paul: concrete syntax provides hub 01:06:17 Grosof: what semantics will PRR have before RIF arrives? 01:07:16 Grosof: semantics expressed at level of meta-model rather than logical conclusions and actions from premises 01:07:33 Grosof: should RIF provide that kind of stuff? finer/deeper/different level 01:07:59 Welty: agree - PRR has different type of semantics 01:08:05 q+ 01:08:11 Barkmeyer: model-theoretic semantics? 01:08:24 Welty: not necessary - but want to ensure same conclusions 01:08:34 Paul: rule vendors won't change engines 01:09:07 Welty: semantics should be specified, allowing rule engine to know how to translate into their system 01:09:58 holger has joined #rif 01:10:22 Paul: requirement is purely compatibility - OMG uses annotated meta-model 01:11:21 q? 01:11:26 ack hak 01:12:24 Hassan: converging - Structured Operational Semantics (SOS) are rules that transform syntactic constructs - formal semantics 01:12:30 http://www.cs.vu.nl/~x/sos.html Structured Operational Semantics 01:12:45 by J.C.M. Baeten and C. Verhoef ? 01:14:05 SOS is a formalism due to Gordon Plotkin 01:14:14 http://portal.acm.org/citation.cfm?id=199472 ? 01:14:58 yes ! 01:15:04 G. Plotkin. A structured approach to operational semantics. Technical Report DAIMI FN-19, Computer Science Department, Aarhus University, 1981. 15 01:15:37 Uli: how many semantics of how many vendors of how many engines? 01:16:33 http://citeseer.ist.psu.edu/plotkin81structural.html 01:16:39 Welty: scope of this WG is primarily its membership 01:17:10 Sandro: charter discussed standard and non-standard (vendor specific) extensions 01:18:12 Welty: non-member rule engine features could be addressed when designing extension mechanisms 01:18:18 Grosof: any Prolog vendors here? 01:18:21 Concerning SOS I have some doubts whether this would cover e.g. non-monotonic logical semantics and other things important for logic-programming based rule languages. This is probably rather a case for research? 01:18:50 Paul: every PRR vendor has features not covered by PRR - subset/least common denominator 01:19:26 Engel: OntologyWorks is Prolog-like 01:19:45 Grosof: nice to interoperate with pure Prolog 01:20:57 Christian: need phase 1 use cases to identify extensibility mechanism requirements 01:21:35 Welty: how to proceed? classification of existing systems? 01:21:51 About phitzler doubts re: SOS and non-motonicity - no problems at all. Read Plotkin's paper (cited above) 01:22:02 Sandro: worked some on "feature matrix" over summer, inspired by workshop 01:23:18 Bijan: could proponents provide (preferably semantic) descriptions of their languages, particularly vendor-unique features 01:23:23 I second Chris Welty's "let's classify systems" initiative. 01:23:34 Bijan: free licenses helpful too :-) 01:24:35 Welty: need some documentation/understanding from vendors 01:24:35 ekw has joined #rif 01:24:57 pfps: user documents generally don't describe semantics - evocative not descriptive 01:26:00 Offtopic: where are the logs of this chat to be found? 01:26:07 Ginsberg: benchmark results may apply here 01:27:07 Bijan: what parts of languages aren't usefully covered by PRR, RuleML, etc.? 01:27:41 Welty: want clear agenda for tomorrow - sense relative satisfaction with classification 01:28:27 Welty: what about use cases? not interchange-centric? higher-level use cases? 01:28:52 Christian: should include test cases from specific applications as use cases 01:29:26 q+ 01:29:33 Dieter: 80/20 rule - RIF needs to support common constructs 01:30:15 hand raised for you, hak 01:30:28 Grosof: use use cases to weight different features 01:31:25 Dieter: strange RDF features to cover PICS use case - over-design 01:32:13 Welty: use cases not carved in stone - don't commit to cover everything 01:32:31 download a beta of Oracle Biz rules at http://www.oracle.com/technology/products/ias/business_rules/index.html 01:32:57 in the .zip file, rl.pdf is a "spec" of the java-like technical rule language 01:34:02 ?: half of use cases can be satisified by web services 01:34:18 ?: when do rules need to be explicitly interchanged 01:34:40 ?: constraint programming engines also important 01:34:45 q+ 01:34:46 ^?^Igor Mozetic^ 01:34:55 q- 01:35:09 Last ?="Philippe Bernard" 01:35:30 Philippe Bonnard 01:35:36 Oh, soory 01:35:50 Gary: 2 types of scenario: interchange (vendor/company/tool), rule use cases 01:36:17 Engel: another scenario - vendor-neutral authoring tool (e.g. Eclipse or SourceForge) 01:37:02 Engel: need to motivate exchange of rules - some engines may not be able to handle even if properly translated 01:37:42 Boley: collaborating rule engines on large distributed problem 01:39:14 Welty: challenge authors of existing use cases, etc. to think more generally and about how they relate to classification 01:40:17 Pellant: won't be here tomorrow - use case: very wide set of behaviors 01:40:22 Welty: please write up 01:41:01 who spoke? 01:41:01 Bijan: don't want interchange language, want Semantic Web Rules Language - others may also - how to accommodate this 01:41:29 Welty: have this conversation at end of tomorrow, after more clear expectations 01:42:10 Welty: view question as largely irrelevant 01:42:30 Welty: interchange already complex, without introducing divide 01:43:10 Welty: can phrase desire as use case or requirement 01:43:44 Kifer: RIF not possible without defining (web) language 01:43:48 I disagree! 01:44:22 Welty: gather requirements based on need to interchange - resulting spec can be viewed as language 01:44:43 Bijan: different between rules language and (semantic) web rules language 01:44:49 ^different^difference 01:45:24 Welty: RIF will be web-savvy interchange format (URIs, interoperability with OWL/RDF) 01:48:09 Barkmeyer: distinguish desire for interchange with what to interchange 01:48:27 Barkmeyer: focus on capabilities 01:49:28 Pascal Hitzler: ambiguous terms between communities: model, etc. 01:50:23 Welty: on-going and evolving process - can adapt as necessary 01:51:41 close of meeting 01:51:46 Deborah has left #rif 01:51:58 start promptly at 9am for "quick" 3 hour roundtable 01:52:02 -hassan 01:52:13 csma has left #rif 01:52:31 msintek has left #rif 02:35:00 disconnecting the lone participant, Meeting_Room, in SW_RIF()11:00AM 02:35:03 SW_RIF()11:00AM has ended 02:35:04 Attendees were JeremyC, +1.650.696.aaaa, +1.604.930.aabb, meeting_room, jeremy, Giorgos_Stamou, hassan, Sandro, PedramAbrari 02:35:35 RRSAgent, generate minutes 02:35:35 I have made the request to generate http://www.w3.org/2005/12/09-rif-minutes.html sandro 03:45:27 Mala_Mehrotra has joined #rif 09:02:21 Zakim has left #rif 16:49:12 RRSAgent has joined #rif 16:49:12 logging to http://www.w3.org/2005/12/09-rif-irc 16:49:12 zakim, this will be rif 16:49:12 ok, sandro; I see SW_RIF()11:00AM scheduled to start 49 minutes ago 16:52:06 hak has joined #rif 16:52:11 msintek has joined #rif 16:53:30 SW_RIF()11:00AM has now started 16:53:37 +hassan 16:56:10 phitzler has joined #rif 16:58:08 -hassan 16:58:10 SW_RIF()11:00AM has ended 16:58:12 Attendees were hassan 16:58:13 Hirtle has joined #rif 16:59:07 SW_RIF()11:00AM has now started 16:59:14 +hassan 16:59:57 MarkusK has joined #rif 17:00:03 Anyone there yet? (Aren't we convening at 9:00 PST?) 17:00:11 +Giorgos_Stamou 17:00:27 -Giorgos_Stamou 17:00:46 + +1.650.347.aaaa 17:02:02 +Giorgos_Stamou 17:02:53 ChrisW has joined #rif 17:02:57 Harold has joined #rif 17:03:03 der has joined #rif 17:03:23 zakim, who is on the phone 17:03:23 I don't understand 'who is on the phone', ChrisW 17:03:28 zakim, who is on the phone? 17:03:28 On the phone I see hassan, +1.650.347.aaaa, Giorgos_Stamou 17:03:37 w3c has joined #rif 17:03:44 zakim, aaaa is meeting_room 17:03:46 +meeting_room; got it 17:04:16 w3c has left #rif 17:05:34 Hallmark has joined #rif 17:05:45 csma has joined #rif 17:06:30 Allen has joined #rif 17:07:26 jeremy has joined #rif 17:09:09 PaulV has joined #rif 17:09:11 mdean has joined #rif 17:09:16 aharth has joined #rif 17:09:38 Donald has joined #rif 17:09:57 Deborah has joined #rif 17:10:43 pfps has joined #rif 17:11:00 vassilis has joined #rif 17:11:31 We are doing *short* use-cases 17:11:46 dlm has joined #rif 17:12:16 bijan has joined #rif 17:12:27 vassilis has joined #rif 17:12:32 holger has joined #rif 17:12:44 jos has joined #rif 17:12:58 Christian: is selling widgets w/o a web presence 17:13:27 Christian: pricing is determined by rules (e.g., discount for large orders or nice customers) 17:14:08 Christian: will accept legal (only) credit cards 17:14:16 zakim, give every speakinger 5 minutes 17:14:16 I don't understand 'give every speakinger 5 minutes', sandro 17:14:21 zakim, give every speaker 5 minutes 17:14:21 I don't understand 'give every speaker 5 minutes', sandro 17:14:29 zakim, give each speaker 5 minutes 17:14:29 ok, sandro 17:15:05 + +1.506.444.aabb 17:15:23 st has joined #rif 17:15:25 Christian: part of the rule (order size) needs information from the web store 17:15:41 zakim, aabb is me 17:15:41 +Hirtle; got it 17:16:05 Christian: part of the rules (customer status) needs information from Christian's local (and private) data 17:16:59 Christian: part of the rules (legal credit cards) needs information from the customer 17:17:06 scribe: pfps-scribe 17:18:46 Sandro: my pet use case has to do with CSCW 17:20:00 Sandro: I use Prolog, and find I need CW stuff 17:20:26 ack John_Hall 17:20:36 q+ John_Hall 17:20:37 ack 17:20:43 ack John_Hall 17:20:54 John Hall: looking at privacy and data protection in EU 17:22:13 John Hall: wants regulations to be in rules, and machine processable 17:23:44 Said: fraud detection use case 17:24:38 q+ Said_Tabet 17:24:39 ack 17:24:56 ack Said_Tabet 17:25:10 Said: problems with privacy, collaboration, etc. 17:25:43 Chris to John: are there extant systems that do anything for you 17:25:53 q+ Uga_Corda, Vassilis_Tzouvaras, Harold_Boley, Jing_Mei, Michael_Kifer 17:26:07 John Hall: vocabulary methods help 17:26:16 ack 17:26:18 John Hall: there are tractability concerns 17:26:26 q? 17:26:32 ack Uga_Corda 17:26:53 http://www.w3.org/2005/rules/wg_f2f_1_participants.html 17:27:19 saidtabet has joined #rif 17:27:38 Ugo Corda: enterprise integration 17:28:18 Ugo Corda: express BPEL stuff in production rules 17:29:14 +??P12 17:29:16 Ugo Corda: BPEL of interest because of transportability 17:29:18 jjc has joined #rif 17:29:44 Ugo Corda: when rules are involved in BPEL, need a transport process for rules 17:29:57 zakim, who is one the phone 17:29:57 I don't understand 'who is one the phone', sandro 17:30:04 zakim, who is on the phone? 17:30:04 On the phone I see hassan, meeting_room, Giorgos_Stamou, Hirtle, ??P12 17:30:32 ack Vassilis_Tzouvaras 17:30:50 Vassilis: BPEL in medical applications require uncertainty handling in both KR and rules 17:31:33 Vassilis: we use a particular fuzzy representation 17:31:38 ack Harold_Boley 17:31:53 Harold Boley: information integration with rules and taxonomies 17:32:16 Talking about http://lists.w3.org/Archives/Public/public-rif-wg/2005Dec/0029 17:32:33 Harold Boley: example is regional business development (done by Canadian federal government) 17:33:14 Harold Boley: several information sources, one with less coverage but better, one with more coverage but not as good 17:33:58 Harold Boley: rules used to determine how to combine the information 17:34:08 Harold Boley: need for URI normalization engine 17:34:43 Harold Boley: fuzzy and uncertainty was also needed 17:36:02 Harold Boley: also co-reference determination, when URIs were not available 17:36:22 Harold Boley: co-reference determination was done by rules 17:36:31 ack Jing_Mei 17:36:45 Talking about http://lists.w3.org/Archives/Public/public-rif-wg/2005Dec/0030 17:37:38 Jing Mei: personal information, particularly preferences (e.g., contact and schedule methods) 17:38:21 ack Michael_Kifer 17:39:07 Michael Kifer: applications evolve, including schema evolution, so language needs to support semi-structured data 17:39:53 Michael Kifer: policies also change, and quickly, particularly if they have internal temporal aspects 17:39:58 q+ Chris_Menzel 17:40:46 Michael Kifer: because of policy change support for rule manipulation (e.g., adding a new rule) is needed 17:41:09 Michael Kifer: one way to do this is reification 17:41:19 q+ "15 Elisa Kendall, Sandpiper" 17:41:26 Chris to Michael: Do you need to get into the rules? 17:41:44 Michael Kifer: no, just describing them from the outside 17:42:02 q+ Elisa_Kendall, Dieter_Fensel, Gary_Hallmark 17:42:25 q+ Joshua_Engel 17:42:51 q+ Holger_Lausen, 17:43:15 Michael Kifer: aggregations of rules may need to be examined, e.g., to extract a rule 17:43:38 Ben Grosof to Michael: aren't there other ways of doing this? 17:44:39 Ben Grosof to Michael: e.g., preferences 17:44:46 We can't hear the questions over the phone !!! 17:45:26 is that better? 17:46:11 Elisa has joined #rif 17:46:11 Michael Kifer: reification is not the only solution, just the most elegant 17:46:57 elance? 17:47:03 elegance ? 17:47:21 ack Chris_Menzel 17:48:38 Chris Menzel: ontology integration (in Boeing) needed more than OWL provided 17:49:19 Chris Menzel: mapping rules are needed between ontologies 17:50:03 Chris Menzel: how to proceed beyond OWL (to rules (SWRL?), to full first-order, to ....) 17:50:36 ack Elisa_Kendall 17:50:50 Elisa Kendall: ODM and OMG need interoperability 17:51:20 Elisa Kendall: want ontologies to express semantic constraints beyond ODM etc 17:51:51 Elisa Kendall: want SW rules to support software generation 17:52:06 http://www.businessrulesforum.com/2005_Product_Derby.pdf 17:52:09 Elisa Kendall: also interested in ontology mapping 17:52:12 ack Gary_Hallmark 17:52:55 Gary Hallmark: business rules 17:53:42 Gary Hallmark: use case - modelling a company in business rules 17:54:39 Gary Hallmark: tools for generating and analyzing and maintaining rules and rule sets, UI tools 17:55:44 -hassan 17:56:37 chrisM has joined #rif 18:13:34