RIF Telecon 31-Mar-09 -- 31 Mar 2009

<ChrisW> scribenick: StellaMitchell

<ChrisW> Scribe: StellaMitchell

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2009Mar/0129.html

<ChrisW> PROPOSED: accept last week's minutes

<ChrisW> RESOLVED: accept last week's minutes

Admin

Liason

<AdrianP> OWL2 has a new overview document

<AdrianP> OWL 2 Web Ontology Language Document Overview

<AdrianP> http://www.w3.org/TR/2009/WD-owl2-overview-20090327/

Chris: liaisons, OWL, functions and predicates

Jos: I think that this is not the right document for the functions and predicates, issue with reusing namespace

ChrisW: You think they should be in DTB then?

Jos: Yes

ChrisW: Does Axel agree?

Jos: haven't talked to him yet about it

Sandro: I'm not sure they should go in DTB
... there isn't a good solution from a user's perspective
... at least if it's in DTB, OWL won't have to pay attention to it

ChrisW: let's see what Axel thinks of this

<ChrisW> ACTION: chris to query axel on whether to move rdf:text F&Os to DTB [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action01]

<trackbot> Created ACTION-719 - Query axel on whether to move rdf:text F&Os to DTB [on Christopher Welty - due 2009-04-07].

Sandro: new document overview published by OWL
... roadmap and high level introduction

ChrisW: we have talked about doing this for RIF also

<AdrianP> we have a little overview of RIF in the UCR document

F2F13

<sandro> http://www.w3.org/TR/owl2-overview/

<ChrisW> http://www.w3.org/2005/rules/wiki/F2F13

<AdrianP> structure of RIF

<AdrianP> http://www.w3.org/2005/rules/wiki/UCR#Structure_of_RIF

ChrisW: draft agenda is available, main objectives are to finalize the working drafts and bring them to last call
... there will be time to work on the documents at the meeting, and hopefully we will vote on last call at the meeting
... we will distinguish between issues to be addressed at the meeting and which will be postponed
... and go through test cases

<ChrisW> registration: http://www.w3.org/2002/09/wbs/38457/f2f13/

<LeoraMorgenstern> I, and probably Stella, can't access that web page

ChrisW: registration/regrets for F2F13 above

I think I will

<LeoraMorgenstern> If I can get a free hotel.

<LeoraMorgenstern> I do have some friends in Boston, so I'll see what I can do.

<ChrisW> great

Action review

ChrisW: action 707 - pending discussion

action 716 - continued

<trackbot> Sorry, couldn't find user - 716

<scribe> ACTION: 714 to continued [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action02]

<trackbot> Sorry, couldn't find user - 714

action 708 is continued

<trackbot> Sorry, couldn't find user - 708

<LeoraMorgenstern> oops

<LeoraMorgenstern> I got disconnected. Will call in again. Hold on.

<LeoraMorgenstern> Short answer: continued.

588 continued

<LeoraMorgenstern> Long answer: I tried to access the issues pages, etc., and can't.

<LeoraMorgenstern> I need to access those pages. Can I get the permissions changed?

ISSUE-67

<LeoraMorgenstern> Okay, I'm back again.

<sandro> issue-67

<sandro> issue-67?

<trackbot> ISSUE-67 -- need string predicates string-less-than, etc. (waiting on PS) -- OPEN

<trackbot> http://www.w3.org/2005/rules/wg/track/issues/67

ChrisW: string less-than predicate, discussed at telecon late last year
... if we have this predicate is would be easier to map "<" in the PS to the predicates

Sandro: multi-typed < comparator

ChrisW: we could then abandon this issue, or do some extra work to find how different existing rule systems handle this

Gary: XPath?

<Michael_Kifer> Prolog has @< for non-numeric comparison.

ChrisW: the current issue is that since RIF is dynamically typed, mapping the "<" in PS to XML may be complicated

<josb> http://www.w3.org/2005/rules/wiki/DTB#pred:matches_.28adapted_from_fn:matches.29

Sandro: and also whether we should have it since XPath dosn't have it

Jos: I think we should drop the redundant items, string-less-than, string-greater-than, string-equal-to

ChrisW: anyone else want to drop them?
... anyone want to keep them? Sandro, Gary

<josb> no

Sandro: I think a lot of people will have to implement the more general predicate anyway

ChrisW: issue-67 has 3 parts. 1. DTB string compare predicates. 2. whether to have general compare predicates, 3....
... Jos, do you feel strongly about part 1, about getting rid of the string specific compares?

Jos: yes
... XPath did a pretty good job of defining operators for comparing, and I think following them is a good idea

Gary: I think XPath is not good in some ways, and that we shouldn't necessarily just adopt it

ChrisW: Gary, would you object to dropping them?

Gary: no

<ChrisW> PROPOSED: drop string<, string> etc. from DTB

ChrisW: would anyone else object to the proposal above (we will vote next week) ?

<ChrisW> ISSUE: Should we have a more general "literal-less-than" (etc.) predicate that covers < tests for all literals.

<trackbot> Created ISSUE-96 - Should we have a more general \"literal-less-than\" (etc.) predicate that covers < tests for all literals. ; please complete additional details at http://www.w3.org/2005/rules/wg/track/issues/96/edit .

<ChrisW> straw poll: -1 object ... +1 would do work to make it happen

<sandro> +1 add literal-less-than, so the PS can have a ">", etc.

<DaveReynolds> +0.1

<josb> -0.9

<ChrisW> +.1

<LeoraMorgenstern> .3

<Michael_Kifer> +0.1

<AdrianP> +0.2

<Gary> 0

<mdean> +0.1

<Harold> +.3

<LeoraMorgenstern> +0.3

jos: reason for not supporting the idea is redundancy and diversion from XPath

Sandro: we have a reason to divert from XPath becase we have a different execution model
... and would be a big benefit in the PS

ChrisW: we resolved that the PS should map directly to the XML

Gary: what happens in RIF if compare two different types of things?

Sandro: undefined truth value, and implementation can raise an error

ChrisW: Sandro, are you willing to edit DTB to define these predicates?

<ChrisW> ACTION: Sandro to add definition of literal-< (etc) to DTB [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action03]

<trackbot> Created ACTION-720 - Add definition of literal-< (etc) to DTB [on Sandro Hawke - due 2009-04-07].

ISSUE-80

<sandro> (where it is NOT yet agreed to keep it.... but there is a SLIGHT WG leaning towards it)

<LeoraMorgenstern> Can you c&p the issue, since I can't access it?

<ChrisW> http://www.w3.org/2005/rules/wiki/DTB#Comparison_for_Literals

ChrisW: I edited DTB to add the predicate that I think we agreed on
... I removed predicate-literal-equals, changed name, used CamelCase, changed description of semantics of LiteralNotIdentical
... see introduction to section 3.1.1

<Harold> http://www.w3.org/2005/rules/wiki/index.php?title=DTB&diff=7950&oldid=7674

Jos: question about use of the term dialect, nt sure it is adequately defined

ChrisW: the wording "dialect at hand" is used throughout the document

DaveR: clarification about disjoint types

<mdean> i have another telecon - bye

<Harold> Span of Chris' changes: http://www.w3.org/2005/rules/wiki/index.php?title=DTB&diff=7961&oldid=7674

<ChrisW> PROPOSED: Accept changes reflected in DTB version [http://www.w3.org/2005/rules/wiki/index.php?title=DTB&oldid=7961], closing issue-80

ChrisW: If we agree to this wording, it will close issue-80

Jos: I had an action to see how this change would affect the OWL-RL ruleset, and I am not convinced that it makes it possible to write a RIF core OWLRL ruleset
... and I think that was the motiviation for this predicate

<josb> l1[differentFrom -> l2]

Jos: RIF core has to be safe, to literals must be introduced in some other place so you can refer so them in the rule body

DaveR: you can write it in the frame syntax

Jos: discussion about literals...will need to take into account all literals in the vocabulary, even ones that aren't mentioned in the ontology

<josb> 1[differentFrom -> 2]

Sandro: I don't see this as a problem

Jos: won't be able have a generic ruleset

ChrisW: does this point hold for the ruleset without this predicate?

Jos: yes

ChrisW: this does reduce the size of the ruleset

<ChrisW> PROPOSED: Accept changes reflected in DTB version [http://www.w3.org/2005/rules/wiki/index.php?title=DTB&oldid=7961], closing issue-80

Jos: I would not object to this

<sandro> +1

<josb> 0

<ChrisW> +1

<Harold> +1

<DaveReynolds> +1

<LeoraMorgenstern> +1

<Michael_Kifer> 0

<Gary> 0

<AdrianP> +1

<ChrisW> RESOLVED: Accept changes reflected in DTB version [http://www.w3.org/2005/rules/wiki/index.php?title=DTB&oldid=7961], closing issue-80

<josb> (assuming the resolution only pertains to the definition of isLiteralNotIdentical))

<ChrisW> ACTION: Chris to close issue-80 [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action04]

<trackbot> Created ACTION-721 - Close issue-80 [on Christopher Welty - due 2009-04-07].

ChrisW: safeness of OWL-RL and Core should be explained somewhere - the fact that you need to ground these literals

<ChrisW> issue-80?

<trackbot> ISSUE-80 -- Should we extend DTB to include more general builtins -- OPEN

<trackbot> http://www.w3.org/2005/rules/wg/track/issues/80

<josb> http://www.w3.org/2005/rules/wiki/SWC#Embedding_RIF-OWL_2_RL_Combinations

Jos: I don't think this text (Chris' comment above) belongs in SWC

Sandro: it should be explained in the embedding appendix

Jos: I wouldn't know what other text to put there

Sandro: how does the embedding compare to the OWL RL ruleset?

Jos: In a RIF-OWL combination where the RIF ruleset is empty, I'm not sure if there would be any entailments that wouldn't be derived from the OWL ontology alone

ChrisW: let's move this discussion to email

<ChrisW> ACTION: Dave to update OWL-RL document to reflect discussion on safeness (esp. in light of new nonidential preciate) [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action05]

<trackbot> Created ACTION-722 - Update OWL-RL document to reflect discussion on safeness (esp. in light of new nonidential preciate) [on Dave Reynolds - due 2009-04-07].

Coreifying SWC

ChrisW: Jos, you raised an issue about making SWC valid for Core. Where do we stand on this?

Jos: I suggest to define the embedding only for the case where it is embeddable into Core.

ChrisW: and this can be checked statically?

Jos: yes

ChrisW: anyone have any questions or concerns about this?

<ChrisW> ACTION: josb to COREify SWC document [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action06]

<trackbot> Created ACTION-723 - COREify SWC document [on Jos de Bruijn - due 2009-04-07].

Jos: wrt coreifying, there is an issue with equality....will need to either have equality, or axiomatize it
... could either define embedding only for BLD, or can restrict to equality free part of OWL2RL
... I prefer to define it only for BLD
... there are many useful OWL2RL statements that use equality, so I'd rather not disallow them in the embedding

<Michael_Kifer> Equality does not add expressiveness only if the arities are bound, like in OWL. In general you need an infinite number of axioms.

ChrisW: for embedding in Core we can have a syntactic restriction....

Sandro: use axioms for equality in core embedding

mk: you cannot axiomatize equality in core

Jos: it would depend on the predicates that are actually used in the ruleset, so it's actually an embedding not an axiomatization

<ChrisW> ACTION: josb to summarize core-ifying owl in email [recorded in http://www.w3.org/2009/03/31-rif-minutes.html#action07]

<trackbot> Created ACTION-724 - Summarize core-ifying owl in email [on Jos de Bruijn - due 2009-04-07].

<ChrisW> aob?

ChrisW: adjourned

- DRAFT -

RIF Telecon 31-Mar-09

31 Mar 2009

Attendees

Contents