Provenance Working Group Teleconference -- 26 May 2011

<trackbot> Date: 26 May 2011

<Luc> Scribe: Tim Lebo

<GK> My VOIP software client license has expired ... I thought it was free for basic use ... will take me a while here to get something sorted.

<Luc> hi tim

<tlebo> scribe: tlebo

<paolo> hallo, has the passcode changed by any chance?

<Luc> no, still same

<Luc> i had to try several times

<paolo> doesn't take it.

<paolo> will try again

<dgarijo> i had to try twice too

<jorn> Zakim: ??p37 is me

#prov, scribe is a newbie, please help me identify the speaker during the discussions.

<paolo> UK number busy and US number won't take my pin.

luc: administrative issues, presentation from incubator, web arch.
... admin issues - approve minutes from last week.

<dgarijo> +1

<khalidbelhajjame> +1

<satya> +1

<smiles> +1

proposed: accept minutes

<frew> +1

<jorn> +1

<pgroth> +1

<jun> +1

<Edoardo_and_David> +1

accepted: minutes

mapping callers to irc IDs - please tell zakim who you are.

<Edoardo_and_David> we might be ??P55

<Christine> I am probably [IP caller]

review ACTION: Tim to add mutli-source. he will add as we go along.

next item - invited experts

Satya's status still has not changed, prov-wg chairs have no insight

<scribe> ACTION: luc to ask w3 about invited experts. [recorded in http://www.w3.org/2011/05/26-prov-minutes.html#action01]

<trackbot> Created ACTION-5 - Ask w3 about invited experts. [on Luc Moreau - due 2011-06-02].

issue: someone please sign up to scribe next week.

<trackbot> Created ISSUE-18 - Someone please sign up to scribe next week. ; please complete additional details at http://www.w3.org/2011/prov/track/issues/18/edit .

next item: participation of discussion. we need others to participate.

<paolo> for the record: I am also requesting an update on how to participate while member status is pending. thaks

<frew> +1

<khalidbelhajjame> +1 for Yolanda idea

luc - interest in presentation to educate membership of group.

<satya> +1 for presentation

<paulo_> +1

<jorn> +1

<Yogesh> +1

<dgarijo> +1

<jcheney> +1

<Edoardo_and_David> +1

<paolo> +1

<CarlOGC> +1

<Christine> A refresher would be useful. Having missed the first couple of meetings of this, I am still trying to get up to speed.

yolanda will present; but what time?

<paolo> 16

<paolo> 4pm UK

<Christine> Ok for me

Monday, 8am Pacific time, 5pm France time

<frew> mon 31 may is a holiday in the us

<dgarijo> i won't be able to make it on monday :(

<khalidbelhajjame> Monday is a bank holiday in the UK

<frew> oops I meant mon 30 may

Tuesday 8am Pacific time

<jcheney> I will be away both days

<paulo_> +1 (Tues)

<Christine> Should be okay

<khalidbelhajjame> +1 for tuesday

<Yogesh> +1

<frew> -1 (conflict)

<jorn> +1 (both ok)

<paolo> I can make it

<smiles> +1

<satya> +1

<Edoardo_and_David> +1

-1 Thursday

<zednik> -1 for Tuesday

<khalidbelhajjame> +1 for thursday

<dgarijo> +1 for thursday

<frew> +1 thu

<paolo> -1 for Thu

<jorn> +1 thursday

<Edoardo_and_David> -1 Thursday

<jcheney> +1

<satya> -1 for thursday (conflict)

<smiles> +1 thursday

<paulo> -1 (Thurs) conflict

<Christine> Thursday is a holiday in Switzerland

<Yogesh> +1

<paolo> +1 for Fri

<satya> +1 friday

<frew> +1 Fri

<paulo> +1 (Fri)

<dgarijo> +1 for friday

<jcheney> +1 Fri

<Yogesh> +1

<zednik> +1 Fri

<Edoardo_and_David> +1 Fri

<khalidbelhajjame> -1 fri

<Christine> - 1 Friday (conflict)

<stain> +1 friday

<smiles> -1 friday (but i don't need to attend)

<jcheney> yes, more important for people who weren't there

<dgarijo> +1 to what paul said

<frew> +1 paul

<pgroth> friday?

+1 friday

<satya> +1

<dgarijo> +1

<Yogesh> +1

<jorn> +1

<VinhNguyen> +1

<pgroth> friday looks good

paulo: the slides can be viewed offline

<dgarijo> if I don't remember wrong, the slides are already available, aren't they?

<dgarijo> Yolanda made a presentation already...

extra call next week 8am Friday for Yolanda to review incubator group.

next item: access and query TF

<jcheney> http://www.w3.org/2005/Incubator/prov/wiki/File:Provenance-XG-Overview.pdf

last week: used section 6 and put on wiki for editing.

<dgarijo> http://www.w3.org/2005/Incubator/prov/wiki/File:Provenance-XG-Overview.pdf

<dgarijo> late , sorry.

graham is not online, olaf also not (both were key participants on email)

luc: we should continue discussion on mailing list; we also should edit the wiki.

<paolo> GK had the same problem

<paolo> persist...

<pgroth> GK it should be ok

+1 mailing list and wiki continuing

<pgroth> did you dial into the us number

<pgroth> ?

simon: good idea, plan for editing wiki page?

how to deal with diverging views? everyone edit?

graham discussed scope - wiki needs to define scope but does not

luc: we should discuss on mailing list and follow up on wiki with results.
... making new pages, commenting - everyone is free to do so.
... remember for f2f, we need proposal and issues against that proposal (for PAQ TF)

next item - issues in system for each definition of the charter concept.

regarding "Resource"

additional ideas for "Resource"

<dgarijo> +q

dgarijo: jun pointed out Resource concept should be broader. We should keep narrower definition of "Resource".

'Web resource"

<martin> +q

martin: resources are digital (e.g. web resource)

-1 for scoping the Resource to digital items.

<satya> It may also be non-digital resource also

<zednik> -1 for treating resources as web resources

martin: Resource State is not clearly defined for digital object; it does not have linear sequence of states; it has versions and derivatives.

<pgroth> +q

RDF working group, "Resource" is conceptual.

paulo ds: [poor voice connection]

paulo ds: do we need an additional definition of resource?

<jcheney> +q

smiles: agree with paulo_ that we should define what we need from "Resource"

<jun> -1 scoping our focus to digital objects

<pgroth> agree with jun

smiles: not good idea to restrict to web resources/ digital resources. derivation: provenance is part of provenance of another. we want to say a digital artifact is the result of a physical process / object.

pgroth: we need to fit with web architecture, which has limits. should our Resource be a broader notion of web's Resource?

<martin> i agree with tlebo

<pgroth> interesting idea

jcheney: what are the consequences of fixing a definition? we can then say what is or what isn't. we should focus on what we NEED to assume about the concepts rather than mapping them all out.

<pgroth> more of a extensional definition

jcheney: list has been saying "we don't care what a Resource is as long as it has a URI"

<martin> this is problematic

<satya> I agree - if we need to describe provenance of something - that is a resource

<satya> so a necessary condition rather than sufficient

satya: an entity of interest to describe provenance - it is a Resource. we are not limiting ourselves with this kind of def.
... regarding compatible with web arch being restricting - what is web architecture, and why be compatible with it?

<Luc> can someone paste the link to the web architecture document?

pgroth: web architecture definitions "Resource" in web context means something particular.

<GK> <cough> *WHY BE COMPATIBLE WITHJ WEB ARCHITECTURE"? This is a w3C WG.

<jcheney> web architecture == http://www.w3.org/TR/webarch/ ?

<Zakim> pgroth, you wanted to respond to satya

pgroth: we should ensure our terms are compatible with web architecture, and if we reuse the terms with different definitions, we need to be clear about it.

<zednik> conceptual model != architecture

satya: what is compatible in web architecture is evolving (e.g. UniProt's RDF wasn't available years ago).

<pgroth> but that's not our remit

just becuase it didn't have a resolvable URI on the web, doesn't mean it's not already a resource (w.r.t the Uniprot's genes, etc)

<khalidbelhajjame> +q

<zednik> +q

jun: just be clear about our terms vs. web architecture's term - we just need a placeholder

<pgroth> we can introduce terms

<jcheney> AWWW sec. 2.2 says: "We do not limit the scope of what might be a resource."

<Luc> ack

+1 on not continuing Resource discussion - w3 gave us this already.

<stain> +1

khalad...: resource state, resource state representation - what do we refer to when we describe provenance - THIS is more helpful. NOT "Resource"

<satya> +1 for khalid's point

martin: we need to distinguish among Resource. information objects behave differently than physical objects. Definition should not be so wide that we don't know what we're talking about. It should be enough to distinguish between physical and digital resource.

<pgroth> but then we then need to use different terms

<smiles> +1 for steffens point

<martin> +1

zednik: conceptual model of provenance might not be the place for Web arch's Resource.

<martin> -1

<pgroth> paulo_ PML uses IdentifiedThing

<GK> @zednik - maybe so, but to do need to know where it touches web resource

<pgroth> +q

paulo_: provenance is in the digital world but is about things in physical world. scientific processes inherently span these.

<pgroth> +q

In the web arch, doesn't 303 redirection handle the digital/physical distinction?

<martin> digital objects can be at different places at the same time. that makes a big diff to physical things

digital objects CANNOT be in different places at the same time. cf FRBR.

pgroth: term Resource in web arch is aligned with what we are talking about.

<jun> +1 pgroth, web architecture allows to make a broad statement about resource

<stain> +1 for yolanda's point of timestamp

YolandaGil: Resource is the hardest to define of the 17. list concepts that are fundamentally different. accessing a UR_L_ at different times returns different digital objects. Let's establish the base of terms we need to discuss (e.g. 7/17)

luc: take this to the mailing list (!)

<martin> +1

<zednik> +1

<smiles> +1

<dgarijo> +1

<jcheney> +1

<paulo_> +1

<Edoardo_and_David> +1

<SamCoppens> +1

<jun> +1

<smiles> OK :)

<Edoardo_and_David> Yes

luc: +1s are on the hook for mailing the list

<GK> I kinda agree with what Yolanda said ... if we leave "resource" wide open, and look at the things we need to express, if a need emerges to say something more specific we'll have a better idea what that is.

<GK> +1 to mailing list :)

[] < who

<GK> Hooked!

<jun> I think Yolanda just said what I wanted to say in a much better way :)

group: the relationships among the terms can reinforce the terms definitions.

martin: e.g. file downloaded from web. at the moment on the server. is it the content of the disk (whcih can be destroyed) or is it the sum of all copies. if you change one copy, how does it affect the rest of it.
... physical - changing it is permanent. and thus linear

luc: we extract notions of Resource from these examples (per Yolanda)

<pgroth> hmmm.... I thought IdentifiedThing == Resource

<pgroth> in PML

<martin> +1

<GK> @martin, difficult to fully understand your point from the IRC log, but I'd say each instance (or item in FRBR terms) is a separate "resource" for provenance purposes.

paulo_: break apart "file" with some properties like where it is located, vs. it is also a "Content container" - need to distinguish content from container.

<martin> @GK sure

paulo_: FRBR's work, expression, manifestation, and item

<pgroth> +q

<dgarijo> a version wouldn't be a state representation of a resource?

paulo_: version is another term related to Resource. Resource cannot be monolithic.

<martin> @tlebo item is the physical level

@martin agreed on physicality of items.

<martin> +1

<satya> @GK: instances may also not be separate resources for provenance purposes - E.g. provenance of all instances of class "human" involves "date of birth"

<zednik> Web Resource State = FRBR Expression

pgroth: we can be guided by charter, but are not constrained by it.

<zednik> Web Resource State Representation == FRBR Manifestation?

<martin> @satya categorical descriptions are even more fuzzy.

paulo_: parts of process hidden that might come up later? e.g. versioning, physical copies.

<GK> @satya Not quite sure your point. Agree that instances may be related, maybe share provenenace. But don't see that means we need to treat every "resource" as complex, stateful thing.

<satya> @martin by categorical description you mean (ontology) class?

martin: museum and ontology for FRBR world. distinction of material object and conceptual level (e.g. Work) - no more than 2(?)

@martin - did you intend to say only 2 levels?

<martin> @satya I mean how a set of items behaves collectively

<YolandaGil> Luc, Paul: Sorry I was not on the IRC before: I can definitely do the Provenance Incubator presentation next Friday at 8am PDT

smiles: mutable things vs immutable things. replicable things vs. unique things. (digital resource copied)

<satya> @gk I agree - how we treat "resource" is mostly driven by requirements of the application in question

<zednik> @martin - FRBR recognizes 4 distinct levels of entities - 3 are conceptual

<pgroth> Thanks Yolanda!

<martin> @zednik discuss by e-mail...

luc: lots of good ideas and points, they need to be expanded on mailing list and wiki to be discussed properly.
... please post to mailing list or wiki.

<satya> @martin ok - behavior of set of items can also be used to group together items?

Yolanda next Friday 8am PDT

<GK> I think FRBR is like ISO 7-layer model. Useful as a way of teasing apart concepts, but not every case where some of the concepts apply neatly divides into all named levels.

bye bye!

<stain> btw - how do you get a name with Zakim?

<GK> Bye. I'll try and get my VOIP client sorted for next time.

<stain> GK, used Blink today (first time) - all right

<Luc> trackbot, end telcon

- DRAFT -

Provenance Working Group Teleconference

26 May 2011

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output