LegalRuleML Inaugural Technical Meeting



LegalRuleML Technical Meeting

Minutes Sept 19th 2012

Time

LegalRuleML TC meeting – Sept. 19th

9,30pm CET

5,30am Australia

3,30pm EDT

12,30pm PDT

Attached document:

1. minute of Sept. 5th

Participants

|Individual |Tara Athan |Voting Member |

|RuleML, Inc. |Harold Boley |Voting Member |

|NICTA |Guido Governatori |Chair |

|University of Bologna-CIRSFID |Monica Palmirani |Chair |

|RuleML, Inc. |Adrian Paschke |Secretary |

Agenda

* Approval of the previous TC minutes

* Update version of the Temporal model

* Glossary

* TOC and schedule for the future

* Proposal to change the day of the TC meeting

* Other business

Monica opened the LegaRuleML TC meeting in time. In total 25 people members. Only 9 people have the voting right. Monica checks who is present: 5 people are in the call. As there is a quorum, the meeting starts. Each participant registered the attendance into the OASIS Kavi system.

The agenda is presented for collecting further amendments or integrations.

No integrations are required by the attendees.

Minutes of the previous TC Meeting

The TC skipped the point to the next TC meeting or by email.

Update version of the Temporal model

Tara presented the metamodel modifications. In particular the session concerning the section concerning the Association, the TemporalCharacteristic.

TOC and schedule for the future

The TC decided to have a reformulation of the TOC according to the existing parts ready and to have a deadline for the Public Review in Novembre.

A new document on the deontic was send by Guido Governatori during the TC meeting.

PROPOSED MOTION by Tara: after suitable internal review (go through all the examples), the example .lrml and source files are moved to the /approved directory. This is not yet a tag, so modifications would still be possible.

•

•

•

•

•

The TC dicussed this fragment and decided to move into block

The TC discusses also on the Boolean value of "efficacy" of a rule. Monica explained that the status Boolean is not correct. If a rule that doesn't say nothing about its efficacy time, it doesn't imply a status known or false. There are constitutive rules for that in any legal system.

PROPOSED MOTION by Monica: to remake the TOC starting from the modules of PDF diagram.

PROPOSAL MOTION: cancel to the TOC the following issues for this Deliverable: semantic of negation, qualification of norm, RIF, XCAML, use cases.

A proposal to add new paragraph in the TOC is approved: 2.3 Metamodel

Constitutive norm and Declarative norm qualification will be included in the future extension using the same mechanism of the strict/defeasible/defeater qualification.

Naf will be included in the defesible part.

Oxygen Graph in the Documentation

Oxygen: check the copyright for the graph.

Glossary

The glossary terminology was discussed by the TC.

Three new terms will be add just in the glossary not as properties to model in the metamodel:

• Law System

• Normative System

• Source of Law

Other business

No request to change the day of the week for the TC. It is confirmed Wednesday.

Adjournment

The conference finished at 4.57pm EDT duration 1:27:38 and it was adjourned with this agenda:

- Metamodel and XML schema udpates, deontic, 3rd Oct., 3.30pm EDT.

Tentative date 26th Sept. for discussing on the glossary and on the examples.

The next meeting will be held Oct 3rd in Skype,

9,30pm CET

5,30am Australia

3,30pm EDT

12,30pm PDT

Annex from the Skype conference call Sept 19th

[19/09/2012 21:25:46] mp: TC meeting in 5 minutes

[19/09/2012 21:28:14] mp:

[19/09/2012 21:30:23] *** Conference call ***

[19/09/2012 21:33:19] mp:

[19/09/2012 21:38:01] mp: * Approval of the previous TC minutes

* Update version of the Temporal model

* Glossary

* TOC and schedule for the future

* Proposal to change the day of the TC meeting

* Other business

[19/09/2012 21:39:31] mp: MOTION: take the time for reading the minutes and vote it by email

[19/09/2012 21:39:46] Tara Athan: +1

[19/09/2012 21:39:48] mp: +1

[19/09/2012 21:39:48] Guido Governatori: +1

[19/09/2012 21:40:10] Tara Athan:

[19/09/2012 21:40:29] Tara Athan:

[19/09/2012 21:50:35] Harold Boley: Discussion about

[19/09/2012 21:55:02] mp: distinction between "applies" and "has"

[19/09/2012 21:56:41 | Edited 22:07:05] Harold Boley: 'External' Temporal:

[19/09/2012 21:57:01 | Edited 22:07:54] Harold Boley: hasStatusDevelopment belongs to TemporalCharacteristic.

Useful for XQuery etc.

'Internal' Temporal:

[19/09/2012 22:04:31] Guido Governatori: p => q

[19/09/2012 22:04:51] Guido Governatori: P:time1 => Q:time2

[19/09/2012 22:06:31] mp: it is great! Tara

[19/09/2012 22:07:40] Adrian Paschke: the external temporal properties in the rule context are more like qualifying metadata of the knowledege in the kb

[19/09/2012 22:09:23 | Edited 22:17:07] Harold Boley: Right, in the "Previous Work" section we should say -- so readers will see that we know the literature -- that

both the database and document-archiving communities have looked into similar external metadata.

[19/09/2012 22:10:26 | Edited 22:33:31] Adrian Paschke: yes, this also known/used in life cycle management of knowledge bases as we have it e.g. in Prova and Reaction RulemL with descriptive and qualifying metadata and metadata scoped reasoning (see also the RBSLA project).

[19/09/2012 22:12:02] mp: working draft document

[19/09/2012 22:12:38] mp: working draft document into Committee Specification

[19/09/2012 22:18:14] mp: Tara: how to maintain the document updated with the SVN repository?

[19/09/2012 22:18:34 | Edited 22:18:59] Harold Boley: Tara: If we 'include' schemas and instances into our document sections, then we don't have the problem of "double uipdating" in case of revisions.

[19/09/2012 22:18:50] Tara Athan: Suggestion: if such "include" is not possible, in working documents, put only hyperlink until XML files are tagged

[19/09/2012 22:21:39 | Edited 22:24:01] Tara Athan: PROPOSED MOTION: after suitable internal review (say two weeks), the example .lrml and source files are moved to the /approved directory. This is not yet a tag, so modifications would still be possible.

[19/09/2012 22:21:49] Guido Governatori: +1

[19/09/2012 22:22:01] mp: +1

[19/09/2012 22:22:16] mp: +1

[19/09/2012 22:22:38] Adrian Paschke: +1

[19/09/2012 22:23:05] mp: PROPOSED MOTION: to remake the TOC starting from the modules of PDF

[19/09/2012 22:23:47 | Edited 22:24:32] Harold Boley: +1 for "after suitable internal ..." (better may be walk-throughs of all .lrml files and schemas)

[19/09/2012 22:24:09] Guido Governatori: +1 (Monica proposed motion)

[19/09/2012 22:24:31] Guido Governatori: +1 (Harold amendment)

[19/09/2012 22:24:47] mp: +1 (Harold amendment)

[19/09/2012 22:25:23 | Edited 22:25:37] Adrian Paschke: yes, we should go through all examples and then approve them one by one. I remember from the test cases in RIF that we discovered many errors when we went through them and approved them one by one.

[19/09/2012 22:28:51] Harold Boley:

[19/09/2012 22:29:21 | Edited 22:29:58] Tara Athan: Keypoints - References, LegalStatements, Associations, RuleContext

Neglected - RuleText, other metadata, no temporal

[19/09/2012 22:30:24] Tara Athan: temporal.lrml

[19/09/2012 22:30:25] mp:

[19/09/2012 22:31:05] Tara Athan: Keypoints- Time (from RuleML), Temporal characteristics (external temporal), and use in Rule context

[19/09/2012 22:31:31 | Edited 22:33:27] Tara Athan:

Keypoints: Rule texts with embedded RuleML, strength (both in RuleText and in RuleContext) and overrides

[19/09/2012 22:31:50] Adrian Paschke: the conneciton dropped?

[19/09/2012 22:32:08] Harold Boley: yes, "There's a problem with this call"

[19/09/2012 22:32:23] Harold Boley: ... "Hold on while we try to get the call back"

[19/09/2012 22:32:35] Harold Boley: Now it dropped entirely.

[19/09/2012 22:32:47] Guido Governatori: all of us

[19/09/2012 22:33:01] Harold Boley: Let me call group.

[19/09/2012 22:33:05] Guido Governatori: ok

[19/09/2012 22:33:07] *** Conference call ***

[19/09/2012 22:34:32 | Edited 22:36:23] Harold Boley: Until Monica is back, we are talking about context examples.

[19/09/2012 22:34:35 | Edited 22:36:20] Tara Athan:

Keypoints: metadata, no encoding of rule texts because depend on deontic logic

[19/09/2012 22:35:21] mp: ok I am

[19/09/2012 22:37:37] Adrian Paschke:

[19/09/2012 22:43:04] mp:

[19/09/2012 22:44:32] Adrian Paschke: so the semantics of rule context is that it always "updates" the properties defined on the rule text?

[19/09/2012 22:55:46 | Edited 22:56:23] mp:

[19/09/2012 22:56:28] Tara Athan: @Adrian - yes, that was my idea about the semantics

[19/09/2012 23:00:21 | Edited 23:00:32] mp: comments on the examples: skype or email

[19/09/2012 23:00:49] Tara Athan: Tara's Homework - move all into

[19/09/2012 23:01:38] mp: Law System

Normative System

Source of Law

[19/09/2012 23:02:37] mp: just for the glossary

[19/09/2012 23:03:07] Adrian Paschke: is it this glossary?



[19/09/2012 23:03:25] mp: efficacy, interval of efficacy, enter in efficacy

[19/09/2012 23:03:44] Tara Athan: to be efficacious

[19/09/2012 23:05:04] Harold Boley: @Adrian, in Monica's 2nd email today, the attachment definitions.doc.

[19/09/2012 23:05:55] Tara Athan: Efficacy, Applicability, ?

[19/09/2012 23:06:23] Tara Athan: A rule has efficacy, a rule is efficacious.

[19/09/2012 23:06:28] Harold Boley: Do we want a 3-valued logic?

[19/09/2012 23:07:28] Harold Boley: efficacy=true|false

or

not mentioned, equivalent to efficacy=unknown

[19/09/2012 23:08:04] mp: if it is not mentioned there are default rules in constitution

[19/09/2012 23:08:20] Tara Athan: Enforceability

[19/09/2012 23:09:53] Adrian Paschke: isn't it a simple qualifying metadata, i.e. name value pair, where the value is true or false and might get updated.

[19/09/2012 23:12:57 | Edited 23:13:32] Tara Athan: At time t1, Rule is efficacious. - true/false/unknown

[19/09/2012 23:15:59 | Edited 23:17:51] Harold Boley: Model Checking Partial State Spaces with ***3-Valued Temporal Logics***:

Combining Interval Arithmetic and ***Three-Valued Temporal Logics*** for the Verification of Analog Systems:

[19/09/2012 23:16:14] Tara Athan: At time t1, rule has efficacy - true/false/unknown

[19/09/2012 23:16:59] Tara Athan: at time t1, rule is in force.

[19/09/2012 23:17:11] Tara Athan: at time t1, rule has ???

[19/09/2012 23:19:48] mp: semantic of negation

[19/09/2012 23:21:41] Guido Governatori: naf A

[19/09/2012 23:21:56 | Edited 23:23:50] mp: PROPOSAL MOTION: cancel to the TOC the following issues for this Deliverable: semantic of negation, qualification of norm, RIF, XCAML, use cases

[19/09/2012 23:22:02] Tara Athan: Tara- imported RuleML schema will be changed to include only Neg

[19/09/2012 23:22:08 | Edited 23:22:25] Adrian Paschke: if we exclude naf, we would also exclude production rule systems as engines for Legal RuleML which typicaly only support inflationary negation (a kind of naf)

[19/09/2012 23:30:28] Adrian Paschke: how does the "full majority vote" work in the public review phase of 30 days?

[19/09/2012 23:31:12 | Edited 23:32:29] Adrian Paschke: does every OASIS member have an appointed voting member who will be asked to vote?

[19/09/2012 23:31:42] mp: constitutive norm and declarative norm

[19/09/2012 23:32:12 | Edited 23:32:41] Harold Boley: With "semantics of negation" Monica seems to mean *legal* semantics of negation.

So, a technical naf construct could still be there, as Guido mentioned, 'implemented' with defeasible logic.

[19/09/2012 23:32:17] Guido Governatori: constitutive and normative

[19/09/2012 23:32:32] mp: @Harold:yes legal semantics of negation

[19/09/2012 23:33:49] Adrian Paschke: I see, so it is a "natural language" negation coming directly from the legal text

[19/09/2012 23:33:59 | Edited 23:35:22] Guido Governatori:

[19/09/2012 23:35:22 | Edited 23:35:34] Harold Boley: FORTH STEP --> FOURTH STEP

[19/09/2012 23:39:02 | Edited 23:45:42] Harold Boley: Maybe make the Metamodel not Normative initially (more didactic), but if we have time we can do it later by formalizing the translation from RDFS to XML.

[19/09/2012 23:40:50 | Edited 23:46:09] Harold Boley: Tara/Adrian: Otherwise we would need to define compliance to RDF.

[19/09/2012 23:46:06] Tara Athan: I think compliance to an RDFS schema would mean - IF you use RDF to serialize LegalRuleML, then you MUST use the concepts introduced in the RDFS schema to do so.

[19/09/2012 23:46:32 | Edited 23:46:44] mp: PROPOSAL: 2.3 Metamodel

[19/09/2012 23:46:49] Guido Governatori: +1

[19/09/2012 23:46:53] Harold Boley: +1

[19/09/2012 23:46:56] mp: +1

[19/09/2012 23:46:56] Adrian Paschke: +1

[19/09/2012 23:46:58] Tara Athan: +1

[19/09/2012 23:47:01] Guido Governatori: Approved

[19/09/2012 23:49:27] mp: Oxygen: check the copyright for the graph

[19/09/2012 23:49:37] Adrian Paschke:

[19/09/2012 23:53:46] Harold Boley: Later: Check if RDFS could be published in a non-XML form like N-triples/N3/Turtle, from which XML can be automatically generated.

[19/09/2012 23:54:13] Adrian Paschke: in RIF we included the XML schema in the recommendation document, e.g. RIF PRD schema as an appendix in the spec ()

[19/09/2012 23:54:38 | Edited 23:55:12] Harold Boley: Yes, publish both 'presentation syntax' and XML.

[19/09/2012 23:54:58] mp: [mercoledì 19 settembre 2012 21:38] mp:

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download