Provenance Working Group Teleconference

22 Dec 2011


See also: IRC log


pgroth, Paolo, Curt_Tilmes, pchampin, davidschaengold, +1.315.330.aabb, tlebo, dgarijo, [IPcaller], +44.238.059.aacc, +1.518.633.aadd
Paul Groth
Paolo Missier


<trackbot> Date: 22 December 2011

<pgroth> Scribe: Paolo Missier

<pgroth> thanks paolo

<pgroth> everything is setup


<pgroth> http://www.w3.org/2011/prov/meeting/2011-12-08

<pgroth> PROPOSED to accept the minutes of the Dec. 08 telecon

<dgarijo> +1

<Curt> 0 (not present)


<tlebo> +1

<davidschaengold> 0 (not present)

<jcheney> +1 minutes

<pgroth> Accepted minutes Dec 8 telecon

<pgroth> http://www.w3.org/2011/prov/meeting/2011-12-15

<pgroth> PROPOSED to accept the minutes of the Dec. 15 telecon

<dgarijo> +1

<Curt> 0 (not present)

<StephenCresswell> +1


<tlebo> +1

<jcheney> +1 minutes

<davidschaengold> 0 (not present)

<pgroth> Accepted minutes of the Dec 15 telecon

Paul: next telecon Jan 5th

Plan for Second Working Drafts

<pgroth> prov-aq

<pgroth> http://lists.w3.org/Archives/Public/public-prov-wg/2011Dec/0232.html

<pgroth> +q

target for resolving issues on prov-aq: one week prior to F2F2

Tlebo: issue of advertising where to find provenance

Pgroth: to be addressed after 2nd WD

Tlebo: why so elaborate? one paragraph should suffice

Pgroth: GK thought it may be too much for this version of the draft.

<pgroth> +q

Tlebo: GK and I have scheduled a discussion for Jan.

Pgroth: anyone from PROV- primer? (no)

<Luc> http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html#changes-since-second-public-working-draft

Pgroth: PROV-DM plans for 3rd WD, asking Luc

Luc: issues just posted have already been worked on
... new section created where all constraints are collected
... this gives constraints more space in the doc
... relations such as "wasComplementOf" reworked, Paolo sent a draft proposal
... also reworked recipe -> plan
... also tried to clarify "identifiers"
... collections still being worked on (Paolo)
... most issues raised in the tracked have been addressed. some will not happen (new example, request from Yolanda)
... debate b/w ids and accounts still needed.
... end of January target for next release (3rd WD)

<pgroth> +q

Luc: vote proposed around Jan 20th

<correct> end of January target for next release VOTE

Luc: need feedback. Will invite WG to review the whole doc by Jan 15th
... issue authors please comment o the state of your issues

<dgarijo> I'm here, but I've been out this week.

pgroth: plans for PROV-O?

Satya: discussed org for next version, but no call this week
... discussed updating constraints part, following updates in PROV-DM.
... no constraints were included in the first release, that will happen in the next
... realigning recipe -> plan
... new complementOf proposal to be discussed in the next telco
... ditto for the data containers proposal (aka collections)

Luc: timetable?

Satya: reorg: Jan 10th target date. New features as discussed: around Jan 15-20
... release of next version: around last week of Jan

Luc: concerned about focus on the constraints -- they may still evolve in PROV_DM depending on discussion outcomes
... lightweight vocabulary for all the PROV-DM is desirable, constraints to be added later after their defs settles
... F2F 2 will talk about interoperability, so encoding of the vocabulary is essential if we are to be able to do that

Satya: substantially agrees that all DM terms are reflected in -O

Pgroth: please send out an update to the list after the next confcall with the new schedule

Satya: will do -- also, all invited to the -O call

<pgroth> q


<pgroth> http://www.w3.org/2011/prov/wiki/FormalSemanticsStrawman

jcheney: rearranged the structure and terminology following -DM
... instance of the DM viewed as a formula
... "objects" with a lifetime, attributes. "objects" captures what is common to entities, activities, agents
... events are instantaneous but they can have attributes as well, so they are objects as well
... (presentation on semantics)
... not clear whether time interpretation is needed. GK suggested to avoid time, which may be possible and simpler
... : some things may hold "at any time" (see complementOf)

others depend on time if the mapping ids- >object is independent of time, then fewer things will depend on time in the model

del/others depend on time

Luc: this helps understanding the DM. what's the best way to discuss about specific points?

jcheney: email the list. use actual issues for specific things
... e.g. the recent complementOf discussion was appropriately on the list

satya: time invariant/time dependent interpretations -- -DM is based mostly on events. are events going to replace time in -SEM?

<pgroth> P5 can you mute?

jcheney: time instants are arbitrary partial order. total order not needed. this maybe close to the notion of events in -DM
... ex: use, generation, activity start and end are events in -DM -- so events are associated with the specific relations and properties used in -DM

<satya> @James- yes, that is something I would like to be considered and discussed

jcheney: there is an assumption that events are instantaneous, however they may span a temporal interval. This would make them similar to activities, possibly those that cannot be decomposed. but this is premature
... need feedback before moving on to additional sections
... complementarity is what it was prior to the latest discussion. It may not necessarily match the draft.
... also added sections on the viewOf/foobar relations and tried to give them a formal meaning
... viewOf anti-symmetric, lifetime containment, and relationship amongst the sets of attributes of the two entities

<pgroth> type it

jcheney: foobar (now changed to alternateOf)
... also being updated
... mismatch between a property like "two entity IDs describe the same entity" vs "two entity IDs are associated with two entity records about the same real world thing, but which give different characterizationss"
... need further discussion over email
... new recent changes in -DM regarding IDs of entity vs entity records should now be taken into account

Luc: in -DM we have 3 levels: things in the world, entities, and entity records. Their distinction is important. In the -SEM doc the same distinction ought to be made.

jcheney: yes that pushes -SEM in the direction of having 3 levels rather than 2 (as it is now in -SEM)

pgroth: discuss identifiers or accounts?

PROV-DM Identifiers

Luc: in -DM entities have identifiers, entity records include the ID of the entity, but we do not identify the entity record /itself/ using that same identifiers

pgroth: id of an entity record -> account + entity Id. what's the implication from an access point of view? there is no URL direclty for the entity record

Luc: there should be no impact on the PAQ

entity records must be searched possibly across multiple accounts. this has always been the understanding of how access works

pgroth: now we only indirectly reference an entity record

<tlebo> pgroth, an "entity record" is a chunk of an RDF abstract graph. An entity is mentioned and described in that graph.

Luc: there was never a direct way to refer to entity records. it was always in the context of a given account

<tlebo> +1 to avoid naming entity records.

Luc: we have never been keen on minting new IDs for every entity record created. Req. was to minimize burden on asserters
... what do the SW people / PROV-O team think of accounts and how they can be mapped to SW constructs (eg named graphs)?

tlebo: accounts are being mapped to NGs, it seems to map to "abstract graph" -- indep. of encoding, location,etc.

<Christine> apologies .. I need to leave .. thank you Paul and Luc

NGs are generally meant as graph containers. there is still some confusion on which interpretation of the NG should be used in this context

<satya> @Luc, sorry I have lost audio, will try to respond by mail to the updated account description in DM (an any points that Tim discusses regarding mapping to named graphs)

will work off the concrete example in the id-accounts section in -DM

Luc: can we focus on all the example of accounts found in the -DM doc?

to see what maps naturally and what doesn't

Luc: account IDs are meant to be global, precisely to align with NG id convention

Tlebo: SW does not provide global URIs for NG. some reconciliation will be needed

Luc: there is no obligation to encode account IDs are URIs

<pgroth> trackbot, end telecon

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2011/12/22 17:04:41 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

No ScribeNick specified.  Guessing ScribeNick: Paolo
Found Scribe: Paolo Missier
Default Present: pgroth, Paolo, Curt_Tilmes, pchampin, davidschaengold, +1.315.330.aabb, tlebo, dgarijo, [IPcaller], +44.238.059.aacc, +1.518.633.aadd
Present: pgroth Paolo Curt_Tilmes pchampin davidschaengold +1.315.330.aabb tlebo dgarijo [IPcaller] +44.238.059.aacc +1.518.633.aadd
Agenda: http://www.w3.org/2011/prov/wiki/Meetings:Telecon2011.12.22
Found Date: 22 Dec 2011
Guessing minutes URL: http://www.w3.org/2011/12/22-prov-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

[End of scribe.perl diagnostic output]