Data on the Web Best Practices Working Group Teleconference

25 Sep 2015


See also: IRC log


Caroline_, deirdrelee, laufer, RiccardoAlbertoni, antoine, makx, jerdeb, phila, newtoncalegari, adrianov, nandana, giancarlo, BernadetteLoscio, PeterWinstanley, Sumit_Purohit_, annette_g, ericstephan, SumitPurohit, Caroline, yaso
Steve, Hadley
laufer, Caroline_, adrianov, PeterWinstsanley, deirdrelee, nandada, yaso


<trackbot> Date: 25 September 2015

<phila> Meeting: DWBP F2F day 2

<deirdrelee> scribe: laufer

deirdrelee: welcome to the participants
... continuing Best practices session
... phil update the data identification session in the document

<Zakim> antoine, you wanted to ask about some removed BP

<antoine> http://lists.w3.org/Archives/Public/public-dwbp-wg/2015Sep/0090.html

antoine: asking about a question sent by e-mail
... a question about the vocabulary issues
... I see not a great problem but want to discuss some requirements

BernadetteLoscio: we need to make a list of the requirements and maybe we need a section in the end of the documento to define waht requirements are necessary or not
... to define what requirements are out of scope of the group

phila: what we did is to point to other groups

BernadetteLoscio: we have to review the list but if we find a req that is not there what we have to do

deirdrelee: we can maybe just identify what best practices point to the requirements

antoine: maybe what we have to do in some cases is just to put some explanation in the document

BernadetteLoscio: what is the best way to do this: a section or a text in the intro

deirdrelee: creating a specific action to that

<phila> ACTION: Bernadette to tabulate requirements against the BPs that address them [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action01]

<trackbot> Created ACTION-205 - Tabulate requirements against the bps that address them [on Bernadette Farias Loscio - due 2015-10-02].

antoine: a comment to the editor that the text was not completely removed form the doc

deirdrelee: http://www.w3.org/2013/dwbp/track/issues/146

Maturity Levels in the BP Doc

<phila> scribe: laufer

BernadetteLoscio: talking a bout the maturity level to replace or not the use of the RFC words
... for the maturity model we do not have yet a real proposal but we have some ideas
... laufer send an initial list of the groups of bps using the words
... then an idea was raise to define levels for the bps

<BernadetteLoscio> https://docs.google.com/a/cin.ufpe.br/spreadsheets/d/1IT6IEeyGUY9crIYY9hDQLgdVx4XVzzKYf21N7YQrO5s/edit?usp=sharing

BernadetteLoscio: I have a other proposal and will put soon a link to that
... the idea was to identify the main aspects of a dataset
... if the bps are really used what be the result
... I tried to idetify some of these characteristics
... bernadette is describing the document

there is a link between the chatacteristic and the BP

scribe: agregating some meaning to the BP

if we have levls we do not need to say if one BP is better than othe... r

if I aplly some BP I have for example a level for a characteristic of acessibility

scribe: one BP coul be associated to more that one characteristic
... I want to have a feedback about this first idea

<Zakim> phila, you wanted to ask whether a BP can be associated with more than one aspect? And to ask if Nic.br has a designer

scribe: the second thing is to decide if this enough to evaluate the BP, if this is suficient to replace the RFC keywords

phila: I like the idea of aspects
... phil asked vagner if he has a designer and vagner aswered yes
... not sure if this idea coul replace the RFC keywords

deirdrelee: asking if there are other opinios

antoine: I this is interesting and I am curious of the comparison with the ideas of the quality dimensions

BernadetteLoscio: I was tinking about this yesterday
... I think there is a relation, I do not know if it is a one to one relation
... but if I understood correctly we do not have a fixed list of dimension

<phila> The dimensions from DQV are 7.1 Statistics

<phila> 7.2 Availability

<phila> 7.3 Processability

<phila> 7.4 Accuracy

<phila> 7.5 Consistency

<phila> 7.6 Relevance

<phila> 7.7 Completeness

<phila> 7.8 Conformance

<phila> 7.9 Credibility

<phila> 7.10 Timeliness

BernadetteLoscio: we can do the exercise to see the relations between the aspects of quality and the aspects of the BPs
... I think woul be great to evaluate the BPs
... I think it is interesting

deirdrelee: in general I think it is a very good idea to make the document understandable
... it will be a nice way to transmit the idea of what the document is

<deirdrelee> laufer: i think it's a good idea, it isn't an instance of a document for the dqv, but what we are defining here is a way to define the quality of publishing, we are defining the dimensions and the things we have in dqv. w

antoine: reaction to one thing bernadette said
... dimension mayve to annotations too,

<nandana> +q to ask if the dimensions can be matched to the ones in http://iso25000.com/index.php/en/iso-25000-standards/iso-25012

<Zakim> nandana, you wanted to ask if the dimensions can be matched to the ones in http://iso25000.com/index.php/en/iso-25000-standards/iso-25012

BernadetteLoscio: just completing what laufer said, is more to evaluate the dataset itself instead of the publishing

nandana: to make a list of comprehensive dimensions

<Zakim> phila, you wanted to talk about ODI Certs

<phila> ODI Certs

phila: this discussion make me to remeber ODI certificates
... ODI refers more to the publisher

<antoine> phila+1

we don't have to look to ODI to complete the list of aspects

<phila> I don't think we need to try and match our dimensions to the ODI Certs as they measure the publishing process, not the dataset

deirdrelee: making a specific proposal to this idea

<Caroline_> +1 to phil

<deirdrelee> draft proposal: Add aspects to the BP document to describe the benefit of each BP

<deirdrelee> laufer: for BernadetteLoscio, I agree that quality is about the dataset, but if we say that having a good level of accessibility you need metadata, you're not talking about the dataset, but the publisher

<deirdrelee> .. it's talking aboutthe way it is being published

<deirdrelee> BernadetteLoscio: i understand, but i think in the end what we are going to evaluate is the dataset itself

<deirdrelee> laufer: but the aspects are related to publishing process

<deirdrelee> ... its the quality of the publishing

<deirdrelee> ... it's the information that can help users to understand the datsaet, parse the datset, etc.

<deirdrelee> deirdrelee: the distiction doesn't really matter

<deirdrelee> laufer: a lot of things we don't have to worry about this

<deirdrelee> BernadetteLoscio: it's not just the quality, it's more than the dataset itself

BernadetteLoscio: what we expect of the dataset and the things are connected

RiccardoAlbertoni: just to say that I agree with laufer
... in the BPs we are talking abou the publishing but I also agrre with antoin that probably this metadat information maybe relate dto annotations or other metrics
... we have to identify these distictions

Giancarlo_Guizzardi: some comments
... if we asy that you are talking about the metadata about the dataset we have to establish a connection between the metadata and the dataset

<deirdrelee> draft proposal: Add aspects to the BP document to describe the benefit of each BP

Giancarlo_Guizzardi: I annot see how to avoid the rlation between the metadata and the dtaset itself

<deirdrelee> PROPOSED: Add aspects to the BP document to describe the benefit of each BP

<yaso> -1

<Makx> -1

<Giancarlo_Guizzardi> Yes. the quality of the metadata influences our appreciation of the quality of the data (e.g., the dataset should be described using SHAREABLE metadata)

<Makx> I had the same question as Phil

<Makx> Benefit is in the why

yaso: I think that finding a benefit is very personal, could be money you do not waste

how we can identify that

scribe: I do not think is necessary

deirdrelee: phil said the why section is pointing the reason

BernadetteLoscio: I think why is more descriptive and is different
... I think is more general

<Caroline_> scribe: Caroline_

laufer: we can list some of the reasons to use the BPs
... some aspects are technical
... we can list the tech aspects
... it is parsable

<scribe> scribe: laufer

<Zakim> phila, you wanted to propose that the Why section of each BP be augmented with an aspect - some sort of icon

phila: Why I think is not right, we have the why section and the inteded outcome
... I do not want to change that, I think is useful to read the document

my problem with the proposal is in how it was written

deirdrelee: if we add this idea, to be careful to not exclude other things

Giancarlo_Guizzardi: it is not the case that all the people will address all the bps all the time

is a way to a criteria to use the bps

yaso: whan I said the document is huge, I do not nedd a benefict section
... I like to first read the exmapls and then read the text

so just to say that sometimes is better to be simple

scribe: I thinks we have examples in the document that can clarify these same idea of the aspects

BernadetteLoscio: we are aware that we need more examples
... phil said what I was in my mind

I think for now we should keep the why and have a proposal for the classification and we see that this thing is a complementary thing or if it is redundanct

deirdrelee: le´s go back to the proposal

<phila> PROPOSED: That icons be added to each BP indicating the relevant aspect(s)

<phila> +1

<Giancarlo_Guizzardi> +1

<antoine> +1


<deirdrelee> +1

<Caroline_> +1

<Gisele> +1

<jerdeb> +1

<yaso> +1

<nandana> +1

<RiccardoAlbertoni> +1

<antoine> laufer: not sure the proposal is clear about the icons

<Makx> +1

<antoine> scribenick: antoine


<BernadetteLoscio> +1

<newtoncalegari> +1

<phila> RESOLVED: That icons be added to each BP indicating the relevant aspect(s)

<phila> RESOLUTION: That icons be added to each BP indicating the relevant aspect(s)

deirdrelee: in the past months we've talked about maturity level
... do the aspects replace the maturity levels?
... or are we keeping them as a separate idea?

BernadetteLoscio: annette has seen the discussion
... we were not clear how to describe the maturity levels
... levels for publisher, dataset, process...
... there's a bit about it in the proposal above
... but it's not clear to me


Giancarlo_Guizzardi: if you think of quality measure the name should be different (?)
... maturity gives the idea of something one can always do

<yaso> https://lists.w3.org/Archives/Public/public-dwbp-wg/2015Sep/0037.html

yaso: I was reading the thread
... with laufer's proposal on MUST/SHOULD
... we have to find some way to stick to this
... to substitute the RFC
... keywords

<Giancarlo_Guizzardi> Ontologically Speaking, maturity sounds to me like a disposition to behave producing something with a certain quality. It does not make much sense to use to refer to an intrinsic characteristic of a product (the dataset)

yaso: It's becoming bigger than it needs to be

deirdrelee: does the thread address this?

<deirdrelee> issue-146

<trackbot> issue-146 -- Which section of a BP should be normative? -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/146

BernadetteLoscio: we agree to use RFC keywords
... but I'm afraid we are not using it in the proper way
... I'd like to know if the way we currently use them is ok
... or if we need to change
... (about the titles of BPs)
... Annette doesn't agree with the use of keywords
... they are for systems
... We need a resolution
... from the group

<deirdrelee> http://www.w3.org/2013/meeting/dwbp/2015-09-04

<phila> Annette's e-mail on this topic

deirdrelee: the discussion was on SHOULD and MUST but there are other keywords (MAY etc)

<BernadetteLoscio> https://lists.w3.org/Archives/Public/public-dwbp-wg/2015May/0077.html

deirdrelee: laufer wanted to check this

phila: I don't entirely mind
... looking at other W3C BPs, none of them use RFC keywords
... we would be unusual
... Hadley would probably want to keep them
... because it's a recommendation
... it's supposed to make statements about what to do

<Zakim> phila, you wanted to express ambivalence

phila: Maybe introducing the aspects would alleviate her concerns

deirdrelee: noone else does it for BP but they do it for other docs

phila: yes as annette says it happens for specifying software
... But we don't have sstrict PASS/FAIL tests
... MobileWeb lists all kind of machine tests

<deirdrelee> Remove the RFC2119 normative terms from the BP Document and instead will create our own levels/maturity-model

phila: but even MobileWeb don't have RDF keywords

deirdrelee: is it ok if we replace 'levels' by 'aspects' in the proposal?

BernadetteLoscio: we will work on the aspects and hope they can be used with the same results

<deirdrelee> Remove the RFC2119 normative terms from the BP Document and instead use aspects icons

<nandana> +1 to antoine. I think they cover two different things.

<phila> +1 to Antoine

<nandana> +q to say within the dimension there could be levels

<nandana> antoine++

<phila> +1 to antoine (again)

antoine: levels have the same function as keywords
... aspects have a different function
... working on aspects is good (probably better) but it won't replace the levels/keywords

<Zakim> nandana, you wanted to say within the dimension there could be levels

BernadetteLoscio: agree

nandana: same thing

laufer: the question was about how to use MUST
... if we have a recommendation with a list of MUST

<nandana> antoine, no probs. You said everything I wanted to say already.

laufer: and someone publsihes data that doesn't conform to one
... what does it mean?
... It's a same as the ODI certificates
... we can't forbid data to be published that doesn't comply with the practice

antoine: RFC gives a clear compliance criteron

Giancarlo_Guizzardi: recommendations would carry different weight
... maybe in relation to the aspects.
... One BP has a MUST for one aspect and a SHOULD for another aspect
... I like the idea of design space
... things are not one-dimensional
... people could comply to one aspect and not another
... so we'd have to relate the sue of keywords to aspects

<phila> PROPOSED: That RFC2119 keywords should be removed from the BP doc

Giancarlo_Guizzardi: and that's complicated

<Giancarlo_Guizzardi> +1

<yaso> ]+1

<Makx> +1

<deirdrelee> +1

<phila> +1

<Gisele> +1

<newtoncalegari> +1


<adrianov> +1

<Caroline_> +1

<nandana> +1

<BernadetteLoscio> +1

<laufer> +1

<jerdeb> +1

<phila> Mr Bradner is upset

<phila> RESOLVED: That RFC2119 keywords should be removed from the BP doc

<phila> RESOLUTION: That RFC2119 keywords should be removed from the BP doc

<phila> close issue-146

<trackbot> Closed issue-146.

<phila> close action-197

<trackbot> Closed action-197.

newtoncalegari: should we remove the words altogether?

deirdrelee: it's not the word just the capitalization

phila: be careful with what reSpec does

Discuss the inclusion of a context section (20 min.)

BernadetteLoscio: there was a lot of discussion about data formats

<Caroline_> http://w3c.github.io/dwbp/bp.html#dataVersioning

<deirdrelee> http://w3c.github.io/dwbp/bp.html#context

BernadetteLoscio: I have tried to explain what is a dataset
... yaso says the doc is huge: is it a good idea to include more?

<Caroline_> s/http://w3c.github.io/dwbp/bp.html#dataVersioning/http://w3c.github.io/dwbp/bp.html#context

BernadetteLoscio: it seemed a good idea to make it more precise what our context is
... i.e what we mean when we put data on the web.
... there are additional things - metadata.
... We have values, and a set of metadata
... we are based on web principles

deirdrelee: you're asking feedback on the text or diagram?

BernadetteLoscio: ont he text, maybe there is redundance with the intro
... it needs to be reviewed
... Steve gave feedback.
... Most important point: do we agree about giving meaning to what we talk about?
... cf comments on the web principles by eric XX

deirdrelee: anything visual is helpful

<phila> s/xx/Wilde/

deirdrelee: we have to be careful about the meaning of things in the diagram
... the general idea about giving context is nice
... I had specific concerns abou the distinctions between metadata, info etc

phila: I'm happy
... this document makes a big effort on readability
... diagrams are useful

deirdrelee: we can review it in the next week, once we agree with the general idea.

<phila> PROPOSED: To retain the context section (which, like all of the doc, is open to review)

<deirdrelee> +1


<RiccardoAlbertoni> +1

<phila> +1

<adrianov> +1

<laufer> +1

<newtoncalegari> +1

<Caroline_> +1

<BernadetteLoscio> +1

<phila> RESOLVED: To retain the context section (which, like all of the doc, is open to review)

<phila> RESOLUTION: To retain the context section (which, like all of the doc, is open to review)

Giancarlo_Guizzardi: 'Aspect' is a bad word

<deirdrelee> +1 to Gisele

Giancarlo_Guizzardi: I would use either quality dimension or crtieria

deirdrelee: maybe we don't have to use the word at all

phila: we are using many words

<RiccardoAlbertoni> +1 to don't use the word at all .. :)

deirdrelee: if we keep a connection with DQV then it makes sense to use dimension

<deirdrelee> https://www.w3.org/2013/dwbp/wiki/F2F_S%C3%A3o_Paulo_-_Agenda_Proposal_DWBP

Discuss the glossary creation (should be a section or another document) and resolve open issues/actions (20 min.)

The Glossary

<phila> action-148?

<trackbot> action-148 -- Yaso Córdova to Include a definition of 'a standard' in the glossary -- due 2015-04-20 -- OPEN

<trackbot> http://www.w3.org/2013/dwbp/track/actions/148

<phila> action-154?

<trackbot> action-154 -- Yaso Córdova to Add the definitions of dataset, vocabulary and metadata in the glossary, noting that the boundaries are not always distinct (or helpful). -- due 2015-04-20 -- OPEN

<trackbot> http://www.w3.org/2013/dwbp/track/actions/154

<phila> action-159?

<trackbot> action-159 -- Christophe Gueret to Write a definition of archiving (and preservation) in the glossary (again!) -- due 2015-04-21 -- OPEN

<trackbot> http://www.w3.org/2013/dwbp/track/actions/159

BernadetteLoscio: is it going to be a separate doc? A note?

<phila> action-174?

<trackbot> action-174 -- Ig Ibert Bittencourt Santana Pinto to Make update to glossary for new data usage terms: citations, data producer, consumer, publisher etc -- due 2015-04-24 -- OPEN

<trackbot> http://www.w3.org/2013/dwbp/track/actions/174

BernadetteLoscio: we have a lot of stuff
... we need to check how we are going to do this

<RiccardoAlbertoni> Could you provide a link to the glossary ?

yaso: we already decided it was going to be a separate doc
... it's on github

<yaso> https://w3c.github.io/dwbp/glossary.html

<yaso> https://github.com/w3c/dwbp

yaso: I have 2/3 actions on adding definitions there
... standard, dataset, vocabulary, metadata
... Ig also had an action to define citation, data producer, publisher...
... christophe already made his changes

BernadetteLoscio: it should be published next to the next BP draft

yaso: yes
... but I can't find the decision
... these actions should be closed
... we have to review the definitions

<phila> Notes that action-174 has been completed

<phila> Also action 159 has been completed

BernadetteLoscio: the work on glossary needs to be done in parallel with the BP doc

phila: how many more defintions are expected to be added?
... Not very many, probably

<yaso> +1 to phila

phila: If it's nearly complete, it can be an annex of the BP doc
... this would make it easy

newtoncalegari: if it's a separate we'll have the same problem as the enrichment doc

phila: if there's a lot of terms then it makes sense to be separate
... yaso said we decided to publish separately

<yaso> I agree that it can be with the document

<phila> antoine: If the glossary exists, we can move some of the text from the main document

phila: but I don't think we need to publish separately

BernadetteLoscio: we can check if there are definitions that are in both glossary and main text
... For the next draft we can move the definitions in the main draft and then review.
... Also judge (seeing the size) if it needs to be a separate note

laufer: has the glossary got terms from the BP doc or the vocs as well?
... if we have terms from DUV/DQV then we need separate docs

BernadetteLoscio: it's just for the BP

Caroline_: it's small

<phila> PROPOSAL: The the terms defined at https://w3c.github.io/dwbp/glossary.html be incorporated into the BP document as an annex


<yaso> +1

<newtoncalegari> +1

<laufer> +1

<deirdrelee> +1

<Gisele> +1

<phila> antoine: I can easily point from the DQV to the BP doc terms if needed

<phila> +1

<BernadetteLoscio> +1

<RiccardoAlbertoni> +1

<phila> ACTION: calegari to incorporate the glossary in the BP document [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action02]

<trackbot> Created ACTION-206 - Incorporate the glossary in the bp document [on Newton Calegari - due 2015-10-02].

<deirdrelee> RESOLVED: The the terms defined at https://w3c.github.io/dwbp/glossary.html be incorporated into the BP document as an annex

phila: yaso have you used specific markup?

yaso: I did

<newtoncalegari> http://philarcher1.github.io/dwbp/bp.html#DataIdentifiers

newtoncalegari: I;ve merged a few changed int he data identification section.

phila: action 148 is not done

<newtoncalegari> action-148

<trackbot> action-148 -- Yaso Córdova to Include a definition of 'a standard' in the glossary -- due 2015-04-20 -- OPEN

<trackbot> http://www.w3.org/2013/dwbp/track/actions/148

<phila> close action-154

<trackbot> Closed action-154.

<phila> close action-159

<trackbot> Closed action-159.

<phila> close action-174

<trackbot> Closed action-174.

BernadetteLoscio: I'd like to discuss the previous topic (?)
... related to Eric Wilde's message - hyperlinks etc
... I've answered his message and we had a lot of discussion
... but I don't know how to handle this
... should we had a best practice on this?
... (ie on links between resources and datasets)

<adrianov> scribe: adrianov

<scribe> scribenick: adrianov

yaso: should wait for annette for discussing other BP issues

BernadetteLoscio: I am not thinking about the APIs

<yaso> https://lists.w3.org/Archives/Public/public-dwbp-comments/2015Aug/0003.html

BernadetteLoscio: It is just to have an idea on how to handle this (API discussion)

<Zakim> phila, you wanted to make a proposal

yaso: this is not a simple issue, and demands more time for discussion

phila: a possible proposal is considering changing the name of the document
... publishing data on the web
... we need someway to be confortable on not talking about some things

<Sumit_Purohit_> someone has to use photoshop and add me in that pic :-)

<vagner> d:)

BernadetteLoscio: the last topic is timetable and there is one open issue which is data enrichment
... there is an open issue raised by annette
... then we can move to timetable and next steps

<deirdrelee> issue-196

<trackbot> issue-196 -- Data enrichment -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/196

BernadetteLoscio: issue to be discussed now is 196
... it is important to remember that data enrichment is gonna to be in a separate note, as a member submission

<phila> W3C Member Submissions

Caroline_: will we keep this on the BP?

yaso: write some use cases about data enrichment, and separate two challenges to be also discussed
... data enrichment is a common procedure which also appears in Web data

Gisele: we looked at the document and we have five challenges and use cases

annette_g: data enrichment is not only on data but also on metadata

phila: close the issue and open an action

<phila> close issue-196

<trackbot> Closed issue-196.

phila: action for gisele

<phila> ACTION: Gisele to modify the Data Enrichment Best Practice to cover data as well as metadata [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action03]

<trackbot> Created ACTION-207 - Modify the data enrichment best practice to cover data as well as metadata [on Gisele Pappa - due 2015-10-02].

BernadetteLoscio: we should start the discussion about the next steps and timetable

<Caroline_> Discuss the DWBP timetable and the next steps

<Caroline_> What should be done to improve the document?

BernadetteLoscio: we are unsure about the dates

<Caroline_> Get ideas for the implementation phase

BernadetteLoscio: a first draft should be finished about the next month

<deirdrelee> bp timetable: https://www.w3.org/2013/dwbp/wiki/BP_Document_3rd_Draft

annette_g: I have a proposal for erick raised by issues

<yaso> +1 to annette_g

BernadetteLoscio: could you right or send message about your proposal?

annette_g: sure

<Caroline_> Long term timetable https://www.w3.org/2013/dwbp/wiki/BP_timetable#Overview

<phila> annette_g: I've written some draft BPs on linkable data that I hope address some of what Erik Wilde was raising and I'll discuss them with him

BernadetteLoscio: we would like to review dates and procedures
... discuss next steps and how to get feedback

deirdrelee: lets look to timetable now

BernadetteLoscio: our proposal is to publish the next draft around 25th october
... one of the phase was removed

phila: we are expected to have one more f2f meeting
... tell us about implementation experiences

<phila> Share-PSI implementayion

phila: not suggesting this is what we have to do, it is just a suggestion
... whether the guidelines they offer are consistent with us
... we need to have some evidence of implementation
... experiences of implementation
... have to specify all the criteria
... have to prove that we did all we have said that would be done

<Sumit_Purohit_> +q

BernadetteLoscio: candidate document on beginning of december

phila: december or january does not make much difference

<BernadetteLoscio> https://www.w3.org/2013/dwbp/wiki/BP_timetable#Overview

<nandana> annette_g, I think you need to give access to the doc.

phila: end of candidate recommendation should be march

BernadetteLoscio: if we go for a candidate recommendation on december we will have another document in the begging of next year

phila: not expecting an organization to implement or follow all BPs
... next version is due to next month some time

BernadetteLoscio: I think we will still have a lot of open issues, and also a lot to write and review
... examples to be included, and quality dimensions, classification, context section, glossary etc.
... review BPs and tests

phila: deadline for the last call version of the document is middle november
... last week of november
... will update the timetable

<annette_g> * all the versions are cool *

<ericstephan> ericstephan present+

<Caroline_> https://www.w3.org/2013/dwbp/wiki/BP_timetable#Overview

Caroline_: phila, when should we have to prove that the test was done, a document?

phila: must provide details on how BPs were implemented
... prove everything is done, and show evidence about the implementation
... some changes is usually required at this point

BernadetteLoscio: what happens during candidate recommendation period?

phila: it is important that you get feedback from the group
... feedback about BP implementations

BernadetteLoscio: imlementations should be done bettwen november and march

Caroline_: phila could you include that officialy?

phila: meet in Zagreb for checking BP implementations

<yaso> seems ok

BernadetteLoscio: happy with the result of the F2F

<Caroline_> https://www.w3.org/2013/dwbp/wiki/F2F_S%C3%A3o_Paulo_-_Agenda_Proposal_DWBP

<Caroline_> Please take a look at Other Open Issues - To be discussed/closed before the F2F or if we have time during the F2F

BernadetteLoscio: we have some open issues. reponsibles should take a look and check for necessary discussions.

<laufer> bye all

<RiccardoAlbertoni> goodbye laufer ..

<nandana> bye laufer !!

deirdrelee: end of the BP session
... moving to DUV

Dataset Usage Vocabulary

<deirdrelee> https://www.w3.org/2013/dwbp/wiki/F2F_S%C3%A3o_Paulo_-_Agenda_Proposal_DUV

<ericstephan> https://docs.google.com/presentation/d/1QaLn70KSJ2cBvTsiS3_H59gUq9XHuDe-oDtoViC37Iw/edit?usp=sharing

<deirdrelee> BernadetteLoscio: let's start with slides

<deirdrelee> scribe: PeterWinstsanley

<PeterWinstanley> ericstephan: one of the challenges with DUV is thinking about data usage

<PeterWinstanley> ...DUV has changed a lot since last F2F

<newtoncalegari> A short URL in case you're going to type https://goo.gl/2lJXaj

<PeterWinstanley> Slide #2 shows the eveolved DUV model

<PeterWinstanley> slide '3 has the major components

<PeterWinstanley> ...core was DCAT :Dataset - when thinking about usage we're thinking about the dataset or distribution

<PeterWinstanley> ...what can operate on the dataset

<PeterWinstanley> ...feedback is looking at community responses

<PeterWinstanley> ...citation is looking at references to the dataset - formal or informal

<PeterWinstanley> ...slide #4 - is how a digital citizen examines a dataset and perhaps collaborates with others using tools that are readily available. This would use something that Sumit_Purohit_ developed to help scientists discover datasets they might not be aware of

<PeterWinstanley> ...I might want to look at particular types of datasets that I want to examine with this tool

<PeterWinstanley> ...http://rdesc.org provides illustrations

<PeterWinstanley> ...the distribution is the actual file that I'm interested in

<PeterWinstanley> ...from a DUV perspective the first thought is describing the dataset (breaking some BP rules, but for the point of illustration) I'm calling the dataset directly (slide#7) but it would be better to have a list of search results (slide#6)

<PeterWinstanley> ... the actual data file, in the case of the DUV, would be described in terms of dataset and distribution.

<PeterWinstanley> ...slide#9 The DUV allows me to describe a file that I can use to manipulate the dataset

<PeterWinstanley> ...slide#10 shows a blank screen and provides a dead end (typical when deling with arbitrary data on the web) so

<PeterWinstanley> ...slide#11 shows the ability to provide feedback giving narrative about what I did and what I needed to note in terms of important info

<PeterWinstanley> ...slide#12 I need to know the credibility of the resource (readme files, technical publications, references etc)

<PeterWinstanley> ...slide#13 shows the use of DUV to reference these materials

<PeterWinstanley> ...slide#14 shows the DUV is able to hold all this background information that I currently keep in my head or a notebook, so it provides a mechanism for enriching search resources

<ericstephan> http://w3c.github.io/dwbp/vocab-du.html

<PeterWinstanley> BernadetteLoscio: There are only a few issues, but we can discuss the model

<PeterWinstanley> ericstephan: walking through the document from top to bottom: one aspect is the use of code from existing application to provide a simple way of showing different RDF serialisations: t=turtle, j=json-ld etc

<phila> The j/t switch appeals to my inner geek

<nandana> +1. it is quite nice. Having example in JSON-LD would be quite appealing to some readers.

<PeterWinstanley> ericstephan: the one thing that concerns me is that we need to ensure that blank nodes etc are fixed and to ensure that we provide a good json-ld rendering

<PeterWinstanley> phila: querying the blank node in prov:association

<PeterWinstanley> ericstephan: something we need to check on

<PeterWinstanley> ...i think it is a reflection of the turtle

<Sumit_Purohit_> +q

<PeterWinstanley> annette_g: wondering - it's possible to return information back to the dataset owner, but by what mechanism? there's a value in annotations, but that depends on an ecosystem that provides a route back to the published

<ericstephan> Mozilla Science Lab, Share PSI, US Department of Energy Advanced Scientific Computing Research (ASRC), Force 11 Citations Implementations, Provenance (IPAW), PROV, Open Annotation vocabulary, Research Data Alliance

<PeterWinstanley> ericstephan: that's an area that needs different communities to input on

<PeterWinstanley> ...we might get input from the communities mentioned

<PeterWinstanley> annette_g: as the job is to develop the vocab we don't need to consider how it might be used, but it is worth considering so that we don't do anything incompatible with future generations

<PeterWinstanley> Sumit_Purohit_: to answer your question: we do expect the publisher to provide some mechanism. in the early days of rdesc we thought about semantic mediawiki pages linked off the dataset info. it goes back to the implementation part. perhaps at the time of candidate phase we'll be able to come up with an implementation

<PeterWinstanley> BernadetteLoscio: we have 2 BP related to feedback

<PeterWinstanley> Giancarlo_Guizzardi: still on feedback, and recalling yesterday's discusision on feedback as quality statements, it would be good to include these. Sometimes it is not too realistic (we can't oblige people) but it would be useful to specify some terms

<PeterWinstanley> ...another point: related to activity and the relation to resource and agent. the idea is to reuse as much as possible, but rfom a modeliing perspective we are looking to reify an agent with the resource used

<Sumit_Purohit_> +q

<PeterWinstanley> ...this is a 1:1 relation so the the activity is the event of usage and this event could be looked at from the agent perspective or from the resouce perspective

<PeterWinstanley> ...it's an overkill to reify all of these

<PeterWinstanley> ericstephan: do you have a recommendataion?

<PeterWinstanley> Giancarlo_Guizzardi: is my assumption of 1:1 relation correct?

<PeterWinstanley> ericstephan: that's right

<PeterWinstanley> Giancarlo_Guizzardi: I would fuse activity, association and usage into a single concept

<PeterWinstanley> ...from a modelling point of view we have made it more complex than it need be

<PeterWinstanley> BernadetteLoscio: the main motivation for this is prov reuse; usage is prov:activity

<PeterWinstanley> ...but this is another option

<PeterWinstanley> Giancarlo_Guizzardi: I am just showing the trade off. re-using prov provides more tuples

<PeterWinstanley> ericstephan: hand crafting examples was not easy. there were pieces I needed to put to gether to provide roles. we didn't want to reinvent, we wanted to reuse.

<PeterWinstanley> ...this is the results of working through that

<nandana> +q to remind about the discussion yesterday about renaming Feedback to UserFeedback to keep DQV and DUV consistent http://www.w3.org/2015/09/24-dwbp-irc#T13-54-45

<PeterWinstanley> ...wokring with prov required extra thinking to use

<Zakim> phila, you wanted to talk about notifications

<PeterWinstanley> phila: practicality/usage: any vocab is requiring tooling

<phila> http://www.w3.org/TR/2015/PR-notifications-20150910/

<phila> http://www.w3.org/TR/activitystreams-core/

<PeterWinstanley> ...possible routes: "notifications" and "activity streams" - tool chains to cover things like 'like' in twitter/facebook

<PeterWinstanley> ...the tool could alert the owner of the data when usage/feedback info is created by a user

<PeterWinstanley> annette_g: Sumit_Purohit_ mentioned the idea of a publisher wanting to gather usage and they would create a wiki: probably not practical; plain text might be more practical.

<PeterWinstanley> ...a better way to get detailed description is where data is re-used to make part of the re-usage do the reporting back.

<PeterWinstanley> ...difference between looking and re-using

<PeterWinstanley> ...counting hits on a dataset is different from determining re-use/re-purposing

<PeterWinstanley> annette_g: all the pieces we have available in DUV should be available for usage as annotations. at present they are restricted to fedeback, but they should also be used with e.g. citation

<PeterWinstanley> Sumit_Purohit_: annette_g point is valid: 3 of us from the vocab team will reply collectively

<PeterWinstanley> ...giancarlo's point about feedback, we've thought about it using citation characterisaiotn features

<PeterWinstanley> ... BernadetteLoscio: can explain

<PeterWinstanley> BernadetteLoscio: this is related to yesterdau's discussion on user feedback in the DQV

<nandana> -q

<PeterWinstanley> ...it should be a subclass of DUV:Feedback, but today we should discuss - is this a single class, or are there different types of feedback

<PeterWinstanley> Giancarlo_Guizzardi: we cannot have feedback as a subclass of quality statement

<antoine> +1

<PeterWinstanley> ...it would be useful to refine a typology of feedback, even it we don't oblige people to use it we give them some optional guids

<Giancarlo_Guizzardi> A reference that might be useful in that regard is: http://link.springer.com/chapter/10.1007%2F978-3-319-12206-9_25#page-1

<PeterWinstanley> ericstephan_: one of the things I mentioned to annette - a powerful concept of hte semantic web is the ability transform graphs in any way we need to : annotations could be developed from anything I mentioned in my presentation; I could also translate anything done in DUV into annotations

<annette_g> :)

<PeterWinstanley> ...the things being asked for could be provided by the open annotations model

<PeterWinstanley> ericstephan_: responding to giancarlo; one powerful thing in the open annotation model is the types of interaction. the editors are open to additions to these properties, so we could develop some complementary effort

<PeterWinstanley> Giancarlo_Guizzardi: this was captured in email in the last few days, but we're still left with working with quality feedback ....(?)

<ericstephan_> Annette can I call you?

<annette_g> okay,

<ericstephan_> no feedback

<RiccardoAlbertoni> i will be here after the lunch

<ericstephan_> great, I'd like to touch base with you

<annette_g> * thanks, Phil *

<annette_g> I need to take off, will catch up in a few hours.

<annette_g> bye for now!

<PeterWinstanley> BernadetteLoscio: I know that we need to have feedback, community discussion; it's needed for recomendataion. after the 2nd draft we didn't have much feedback. maybe we didn't search for feedback. what can be done to improve?

<PeterWinstanley> ...this goes for the 3 docs, but specifically for the BP we need implementations and I'd like to discuss how we are going to find/realise these implementations

<PeterWinstanley> ...for feedback, eric can provide a template for this

<PeterWinstanley> phila: talk, conference, tweet, write to other groups, just work on it

<ericstephan_> I think we need to help imagine people how to use it, would be most helpful.

<PeterWinstanley> BernadetteLoscio: other groups: we need help in identifying them

<antoine> This is what I've done last time for DQV: http://www.w3.org/2013/dwbp/track/actions/187

<antoine> in fact it's this: http://lists.w3.org/Archives/Public/public-dwbp-wg/2015Aug/0069.html

<PeterWinstanley> phila: the obvious ones: annotation, geospatial, csv, internationalisation, privacy,... we need to write formally to those we said we'd discuss with

<PeterWinstanley> deirdrelee: not only W3C, but *all* groups

<antoine> but the feedback resulting from my action was not huge :-(

<PeterWinstanley> ...also through work at a national level through our own work

<ericstephan_> The SIOC is another example of who we are working with http://rdfs.org/sioc/spec/

<PeterWinstanley> ericstephan_: i was surprised/delighted by feedback from the SIOC group - but they were asking for an update. we need to branch out into other groups

<BernadetteLoscio> Thanks Antoine!

<PeterWinstanley> deirdrelee: I'll be in Galway in 2-3/52 so will be in contact wiht John Breslin etc

<PeterWinstanley> ericstephan_: it would be good to have a telecon to get their thoughts

<PeterWinstanley> deirdrelee: also check forthcoming conferences

<Sumit_Purohit_> +q

<PeterWinstanley> ericstephan_: sumit and I put in a proposal re: data usage. in the rdesc project, in the DoE advanced computing, rdesc was the only 'semantic web' one, but recently I was in Rockville MD where there were more projects being mentioned but there is a lack of awareness of wider vocabs and a hunger to find out more info

<PeterWinstanley> Sumit_Purohit_: eric and I are proposing one implementation - rdesc is close to what the group has been working on

<PeterWinstanley> deirdrelee: at the stage that editors want feedback if they can create a template email then it will make it easier for mailing list recipients

<Zakim> phila, you wanted to ask about Notes and Recs

<PeterWinstanley> phila: we should send emails, but only after the next formal WDs are pblished

<PeterWinstanley> ...getting one ref implemtation is fantastic, but are we putting them on the REC track? if there are implementiaot reports then it makes the work more credible

<PeterWinstanley> ...can the high bar be met?

<PeterWinstanley> ericstephan_: right now I'm thinking 'notes' but we could show / promote a cohesive standpoint of dwbp group; our work on the 2 vocabs made sense, but we need to review our group cohesion and this will strengthen the 2 vocabs

<PeterWinstanley> antoine: we have to be realistic

<RiccardoAlbertoni> I agree with antoine we have to stay realistic..

<PeterWinstanley> ericstephan_: we've been looking at DUV and I feel good about feedback and citation. DUV was always a little controversial, but at what point is the WG likely to say that we focus more on feedback etc. How can we get confirmation at this point of the groups backing for DUV?

<PeterWinstanley> phila: I have no doubt it is important, not just annotations and citations; I heard giancarlo discussing the problems in incorporating PROV, but essentially I want to see this embedded into CKAN etc

<PeterWinstanley> ...to me the DUV looks on track; the vocabs are late according to the charter, but hopefully after today all 3 vocabs will be on track

<ericstephan_> thank you phil :-)

<phila> == Lunch ==

<jerdeb> i need to leave as well guys

<nandana> bye Makx !!

<jerdeb> have a nice weekend

<ericstephan_> on mute standby music playing in office

<deirdrelee> let's get started!

<ericstephan> I'm here and I think on mute :-)

<deirdrelee> great

<ericstephan> Is anyone speaking?

<deirdrelee> phila: heard back from host in croatia, said it's fine to host us there on March 14th

<deirdrelee> Caroline: is the date definitie?

<deirdrelee> phila: it would be difficult as this is the date that share-psi group have already agreed on

<deirdrelee> ... it would be difficult to change

<deirdrelee> chair: yaso

<yaso> https://www.w3.org/2013/dwbp/wiki/F2F_S%C3%A3o_Paulo_-_Agenda_Proposal_DUV

<deirdrelee> scribe: deirdrelee

<yaso> https://docs.google.com/presentation/d/1QaLn70KSJ2cBvTsiS3_H59gUq9XHuDe-oDtoViC37Iw/edit#slide=id.p16

yaso: we were discussing duv, at the feedback part

BernadetteLoscio: eric, what should we do now?
... should we continue with document or go to model and open issues

ericstephan: since we discussedthe model, maybe we should go to open issues
... we only have a few


<trackbot> issue-178 -- The definition of duv:Feedback needs to be reviewed because it is not clear if it should be a subclass of oa:Annotation or just an instance of oa:Motivation. -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/178

BernadetteLoscio: issue related to feedback

<BernadetteLoscio> ISSUE-173: The use of cito:CitationAct and duv:Citation should be reviewed.

<trackbot> Notes added to ISSUE-173 The use of cito:CitationAct and duv:Citation should be reviewed..

BernadetteLoscio: there are other issues opened by Giancarlo_Guizzardi, they are not on agenda, but they are open issues

<BernadetteLoscio> ISSUE-176: Should prov:SoftwareAgent be used instead of Application/WebOfThing?

<trackbot> Notes added to ISSUE-176 Should prov:SoftwareAgent be used instead of Application/WebOfThing?.

BernadetteLoscio: comments about the last version of the model

<BernadetteLoscio> http://www.w3.org/2013/dwbp/track/issues/177

BernadetteLoscio: issue-173


<trackbot> issue-73 -- What exactly is the audience for the bp doc? -- closed

<trackbot> http://www.w3.org/2013/dwbp/track/issues/73

scribe: confusion about citation act and citantion
... property cite was removed
... i'm not saying the version we have now is the best, but i think this issue can be closed

<yaso> close issue-173

<trackbot> Closed issue-173.


<trackbot> issue-76 -- What advice do we give about publishing metadata so that we identify the intended outcome without making assumptions that maybe false? -- closed

<trackbot> http://www.w3.org/2013/dwbp/track/issues/76


<trackbot> issue-176 -- Should prov:SoftwareAgent be used instead of Application/WebOfThing? -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/176

<yaso> close issue-176

<trackbot> Closed issue-176.

scribe: classweb of things was removed from the model


<trackbot> issue-177 -- Should duv:consumes be used instead of duv:consumed? Should we be able to reify Consumption? -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/177

scribe: there was a property consumes, but this was removed

Giancarlo_Guizzardi_: this is now addressed by the reificaiton of usage

ericstephan: on issue-177, this is where we were considering reusing prov, so i think we can close it, as we've moved on

<yaso> close issue-177

ericstephan: i'm comfortable with closing it, but maybe something that we revisit. maybe to simplify what we're doing, along the lines we were discussing earlier on

<trackbot> Closed issue-177.

BernadetteLoscio: there was another issue related to feedback, which is still open.
... as ericstephan presented, there are 3 parts in the model: usage, feedback and citation

<phila> issue-178?

<trackbot> issue-178 -- The definition of duv:Feedback needs to be reviewed because it is not clear if it should be a subclass of oa:Annotation or just an instance of oa:Motivation. -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/178

BernadetteLoscio: we have to decide what we're going to do, this relates to yesterday's discussion on feedback property that's common with th edqv

<nandana> +q to remind about the discussion yesterday about renaming Feedback to UserFeedback to keep DQV and DUV consistent http://www.w3.org/2015/09/24-dwbp-irc#T13-54-45

BernadetteLoscio: i'd like to discuss all parts of the model, because we have some doubts about the general model, especially the feedback section
... if we have a specific class for feedback, then therre will be a specific class for quality
... should we just have one class for feedback and properties to describe the differnet types of feedback

ericstephan: i liked the discussion on dqv about looking for examples, from the external community
... these things are difficult to do if we don't have examples

BernadetteLoscio: that would be nice,

<Zakim> nandana, you wanted to remind about the discussion yesterday about renaming Feedback to UserFeedback to keep DQV and DUV consistent http://www.w3.org/2015/09/24-dwbp-irc#T13-54-45

SumitPurohit: yesterday we had a resolution to have a feedback a subclass of userfeedback
... i'd like to know if this was a good proposal or not

<phila> Yesterday's resolution was rename DQV:UserFeedback with dqv:QualityUserFeedback making it as duv:Feedback subclass

<ericstephan> +1 nandana

BernadetteLoscio: yesterday we had a discusssion whether to create a subcalss or not, today we should come back and see if thist fits for both vocabulaires

<ericstephan> I like the UserFeedback

<nandana> also to see if it makes to rename the Feedback to UserFeedback (without changing the semantics) in DUV

BernadetteLoscio: we need to decide if we're going to have qualityuserfeedback as subclass of duv:feedbavck


<SumitPurohit> +1 phil

phila: yesterday, we agreed we would

Giancarlo_Guizzardi_: there are 2 issues here
... there is annotation that has 2 possible values that can be assigned to motivation
... we might want to extend this as properties for motivation
... every time we have a type that is restricted to one type of property. from a modelling point of view, there are subtypes that are already there
... this would be a nice way if we want to restrict a subtype of properties that are a subtypes of motivation that refer to quality statements
... from a modelling point of view i don't see any issue wiht this
... we have to check if the values for motivation are enough, if yes
... i think we could have everything. we could specialise user feedback for all of the use cases that are applicable for us

<nandana> +1 to Giancarlo_Guizzardi_

<ericstephan> http://www.w3.org/ns/oa#d4e555 link to motivation

BernadetteLoscio: looking at instance of motivation

<BernadetteLoscio> http://www.openannotation.org/spec/core/core.html#Motivations

BernadetteLoscio: it would be possible if you could send us a model with your suggestion, cos you were in the discussion yesterday too

<antoine> For the record it's better to look at the work in progress at http://www.w3.org/TR/annotation-model/#motivations not the old OA drafts

<SumitPurohit> +1

<ericstephan> +1

BernadetteLoscio: ericstephan, what do you think?
... Giancarlo_Guizzardi_ has an idea to model the specialisation of feedback that will help link the two types of feedback

<Zakim> RiccardoAlbertoni, you wanted to complent the nandana question questions and speak about the modelling

<newtoncalegari> wonders if the work of web annotation wg could help in the issue of annotation: http://www.w3.org/annotation/ http://www.w3.org/TR/2014/WD-annotation-model-20141211/

RiccardoAlbertoni: coming back to nandana's comment, it's not clear whether feedback is just human feedback of also machine feedback
... we might need a new definiton

phila: the annotation group is open to comments

<phila> Motivations

phila: there's a list of motivations
... phila lists all

<yaso> +1 to phila

<ericstephan> I had

<ericstephan> ....

<ericstephan> Yes they seemed to be open to it

phila: antoine said we could define one more, quality user feedback
... or ask them to add one more
... we could add them to create a new motivation for us, and what would they be?/

<Zakim> phila, you wanted to talk about the Annotations WG http://www.w3.org/TR/annotation-model/#motivations

<nandana> RiccardoAlbertoni, I guess in that case we will have the same issue in DQV (dqv:QualityUserFeedback). If they are machines (or any agent), we can't call them users?

<ericstephan> I see significant reuse

phila: providing quality feedback is one, is usage feedback another? is is significantly different to what they have?

anitoine: i am familiar with this work and see a can of worms with asking them to add a new motivation
... it is quite unclear that providing qulity feedback will fall into one of their categories
... what they've tried wiht their current spec is different, more a cross-domain goal
... this might be tricky. however we may still contact them and let them know we want to create a motivation of our own
... but wouldn't push them too much

Giancarlo_Guiz zardi_: agree with antoine, doesn't seem to be related to what we define as feedback, completely differnet

scribe: which begs the quesiton if feedback should be an annotation at all

ericstephan: found that their model is very thorough, annotation seemed to be closesfeedback. to be able to add a comment and reeply back, it seemed like it was applicable from a data usage perspective. maybe not from data quality?t to

<RiccardoAlbertoni> that was my point :)

antoine: agree with ericstephan, from our perspective annotation is important to include. no problem with adding multple annotations on one motivation

<Zakim> RiccardoAlbertoni, you wanted to say that you might associate more than one motivation to an annotation ...

yaso: should we try and reach out to the annotation group still?

ericstephan: before contacting the annotation group, we should have some suggetions. that we have a united front as a WG, for issues that overlap with DQV and DUV

antoine: it's better to clean our act and clarify our relationship between vocabularies, see if we only need one motivation, not so sure
... and then contact annoation working group
... it would be agreat opportunity to get feedback on our vocabs, especially if we give them an excuse by using some of their motivations

<Zakim> phila, you wanted to suggest that we define first and seek inclusion second

antoine: just a matter of timing. once we have made progress and are in a position to get feedback

phila: +1 to antoine I suggest that each/one vocab defines these things
... in the document i suggest we add text to flag that this ns might be moving to another ns

<ericstephan> two action items perhaps for each working group to come up with motivations to discuss internally?

antoine: write an action that takes effect in a couple of weeks, for the editors to contact annotation wg

BernadetteLoscio: i agree with proposal of Giancarlo_Guizzardi_ Let's see what will happen

ericstephan: should we create an action for each vocab
... it can be two actions

phila: when should the action be for?

antoine: second week in october?

<phila> ACTION: antoine to contact OA WG to see whether they would consider adding DQV motivation - due 16 October [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action04]

<trackbot> Created ACTION-208 - Contact oa wg to see whether they would consider adding dqv motivation [on Antoine Isaac - due 2015-10-16].

<phila> ACTION: ericstephan to contact OA WG to see whether they would consider adding DUV motivation - due 16 October [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action05]

<trackbot> Created ACTION-209 - Contact oa wg to see whether they would consider adding duv motivation [on Eric Stephan - due 2015-10-16].

<Zakim> antoine, you wanted to say that I'm leaving

<BernadetteLoscio> thanks a lot!

<RiccardoAlbertoni> bye antoine !

<nandana> bye antoine !!

<yaso> bye, thank you antoine :-)

<phila> issue-178?

<yaso> ISSUE-178

<trackbot> issue-178 -- The definition of duv:Feedback needs to be reviewed because it is not clear if it should be a subclass of oa:Annotation or just an instance of oa:Motivation. -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/178

<trackbot> ISSUE-178 -- The definition of duv:Feedback needs to be reviewed because it is not clear if it should be a subclass of oa:Annotation or just an instance of oa:Motivation. -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/178

BernadetteLoscio: I think we can close issue, because now we have an action

yaso: we can close it? because the definition still has to be reviewed

<yaso> close ISSUE-178

<trackbot> Closed ISSUE-178.

phila: action items arising from it,so it's not being lost

<yaso> http://www.w3.org/2013/dwbp/track/issues/179

<yaso> issue-179

<trackbot> issue-179 -- The Working Group is considering to put all new classes and properties (together with the ones of the Data Usage Vocabulary) in the DCAT namespace. -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/179

phila: i have a view, i think we should put them inthe dcat ns
... dqv starts by saying 'we are extending dcat'
... duv refers to dcat in the intro
... would like to see it in the dcat ns, but open

BernadetteLoscio: both models have dcat,for duv it is core
... how does this happen, we just use dcat?

phila: yes, as long as we don't edit anything that's there
... there is no such thing as the dcat wg, f there is a successor to dcat wg it's this wg

ericstephan: i'm looking at duvmodel, we have things like duv:entity,inherited from prov:entity
... what i like with using dcat ns is that we can guard against overlapping terms in differnet namespaces
... i would be for using dcat ns
... could we make decision just for duv

<ericstephan> it actually makes more sense to have dcat:Entity in a sense.

<nandana> phila, is it possible to edit the text in http://www.w3.org/ns/dcat# to add a link to the DUV document? I guess, yes, right?

BernadetteLoscio: should usage apply to both duv and dqv. but this term is modelled differently in 2 vocabs
... the two vocabs should be consistent in how they model classes, especially if we are going for a common ns

<phila> Yes, nandana, we can edit the /ns doc (but not /TR/vocab-dcat/)

BernadetteLoscio: don't know if this is the best way to model, because in dcat there is only concept of dataset and distribution

<Zakim> RiccardoAlbertoni, you wanted to not sure to understand why to reuse the dcat namespace is such a good idea..

BernadetteLoscio: they don't have a supercall,but in duv we need it, because i don't know how you're going to give properties on both

<nandana> phila, thanks.

RiccardoAlbertoni: what is the justification that using dcat ns is a good idea?

phila: it's one less ns to have to remember. it also makes it more explicit that we're extending dcat for the vocabs, and if feels more coordinated

RiccardoAlbertoni: but in the end we are not going to change the documentation at the ns

phila: right now there is no documentation at the ns, just a link to the ttl file. the definition documnet is separate from the ns doc
... we're talking about adding two new documentation files, not editing the ns doc

RiccardoAlbertoni: as a linked data person, i don't like this
... are we not causing confusion over what part of the model is standard,and what is defined by our group
... dcat is a rec, how will people know what the split is

+1 to RiccardoAlbertoni

<nandana> +1 RiccardoAlbertoni. May be mixing the Rec with Notes might not be a good idea.

scribe: confusing about what terms are the rec standard, and what one are coming from us

ericstephan: if there are no rules in place for reusing an existing ns, and we have a close association with the dcat vocabulary,i don'tsee a problem iwth applying open world asumption to namespaces and using them
... if it will make the vocabs closer to the core
... seems like there's a proprietary nature to namespaces

BernadetteLoscio: i like the idea, but not sure if we can do this now
... see if there is more a maturity model, not sure if we're mature enough to decide this

ericstephan: why don't we do this, and get feedback - myabe we'll get more feedback then

<SumitPurohit> +1

<SumitPurohit> +1 eric

ericstephan: it would be interesting to see how external revieweres would feel about this

BernadetteLoscio: yes, for next draft, we'll say it's an extension of dcat and get feedback

<phila> PROPOSED: That both DQV and DUV documents seek specific feedback on whether or not the terms should be defined in the DCAT namespace

ericstephan: people might get offended :)

<ericstephan> +1

<ericstephan> but is it fair that Antoine is not here? Should the proposal just be for DUV?

<RiccardoAlbertoni> first to have the feedback and than move the namespace

<nandana> +1 for seeking specific feedback

RiccardoAlbertoni: first ask for feedback, then movens

<RiccardoAlbertoni> +1

<WagnerMeiraJr> Bye guys. I learned a lot in this F2F.

Giancarlo_Guizzardi: side issue, just produced model he was discussing, how to send

deirdrelee: put it on the wiki, that's what it's for!

Giancarlo_Guizzardi: idea is that everyone can have a look at it and discuss today

<Gisele> the InWeb team is leaving

<Gisele> thanks for the productive meeting, see u next time

<SumitPurohit> Need to leave for a short meeting.....Will join in half hour (hopefully)

<ericstephan> +1 Deirdre

<ericstephan> No tomatoes

<nandana> +1 deirdre

<Giancarlo_Guizzardi> Here is sketchy visual representation of what I was talking about: https://dl.dropboxusercontent.com/u/14917034/DUV.png

<Giancarlo_Guizzardi> if people can take a look just to check if I got the message across, that would be very nice

<nandana> bye InWeb team !!

phila: feels messy

deirdrelee: it's a/b testing

RiccardoAlbertoni: for sure doing this will have some reaction
... feels it is a bit messy, possible way to go

ericstephan: i like the idea ofcharging ahead a little bit
... unless we try something out and get people's reaction, it will be hard to get people's reaction
... right now we do have two distinct vocabs,
... does anyone know of precedant where a third-party vocab uses dcat

phila: depends if we see our vocabs as addition to dcat, or if it's aligned to dcat

<RiccardoAlbertoni> no.. I don't know an example of vocabulary doing this.. we have the chance to be the first..:)

<ericstephan> +1

<phila> PROPOSED: That DUV begins to use the DCAT namespace, that DQV does not, but that both highlight this as an open issue that will lead to a common way forward in future.

<BernadetteLoscio> +1

deirdrelee: it will def be a way to get feedback from dcat contributors, which is good


<nandana> +1

<phila> +1 expecting some fireworks

<phila> RESOLVED: That DUV begins to use the DCAT namespace, that DQV does not, but that both highlight this as an open issue that will lead to a common way forward in future.

<phila> RESOLUTION: That DUV begins to use the DCAT namespace, that DQV does not, but that both highlight this as an open issue that will lead to a common way forward in future.

scribe: think it's a good idea to go ahead for this draft, but overall not sure if it's a good idea to adopt dcat ns because of rec vs. note issue highlighted by RiccardoAlbertoni

<phila> close issue-179

<trackbot> Closed issue-179.

<phila> scribe: nandada

<phila> scribeNick: nandana

BernadetteLoscio: I would like to discuss each part of the DUV model
... it is good to have a super class for Dataset and Distribution

<yaso> Thank you for participating, RiccardoAlbertoni :-)

ericstephan: I like the idea of reusing DCAT namespace because it relates the vocabs better to their purpose

phila: can we just use prov:Entity instead of duv:Entity?

ericstephan: prov:Entity is too general. That was the motivation to define duv:Entity

BernadetteLoscio: we are proposing to reuse provenance and model as things as activities

<phila> Giancarlo's proposed model for usage

BernadetteLoscio: it's good to use prov and have activity descriptions or should we define our own simpler model?
... what are the motivations for each approach?

Giancarlo_Guizzardi: Using the Usage and Activity as the same might lead to inconsistencies
... in addition to the additional complexity

ericstephan, can you please put that in the chat

ericstephan, I missed it

phila: DQV and DUV looks very similar
... main difference of the new diagram is getting rid of the prov classes

Giancarlo_Guizzardi: describes the proposed vocabulary diagram of DUV
... if QualityUserFeedback belongs to DUV, all types of Feedback should belong to DUV
... we can define subtypes of QualityUserFeedback by defining them using onProperty

BernadetteLoscio: dqv:QualityAnnotation is a subclass of oa:Annotation

Giancarlo_Guizzardi: we are not saying all quality annotation is UserFeedback

BernadetteLoscio: when we see two vocabs are connected, it makes sense to have them in the same namespace

ericstephan: I only have problem with the UserRequestClarification and UserSugestionForCorrection

Giancarlo_Guizzardi: I don't have any problem removing them

<ericstephan> (have to step away for a moment...will be back in two minutes)

RiccardoAlbertoni: does dqv:DataUserRating belong to DUV?
... it was deleted from DQV

Giancarlo_Guizzardi: it naturally belongs to DQV

RiccardoAlbertoni: these deleted classes are present in DUV examples

<Giancarlo_Guizzardi> Just to correct one of Phil's previous statements: https://dl.dropboxusercontent.com/u/14917034/DUV.png Giancarlo's proposed model for usage

BernadetteLoscio: this was an initial proposal. we need to work on this.
... connecting to the two vocabularies seems a good idea

<Giancarlo_Guizzardi> it should be read as Giancarlo's (sketchy) model for Feedback

BernadetteLoscio: should we specialize the Feedback class?

<ericstephan> +1 phila not enumerating all the types of feedback.

phila: we can't enumerate all the possibilities of feedback

<ericstephan> Sumit could you look at https://dl.dropboxusercontent.com/u/14917034/DUV.png and give your reaction?

phila: i don't see the value of classifying feedback.

Giancarlo_Guizzardi: these subclasses don't have to be disjoint
... we can classify ones that are directly related to quality

phila: what about the way Github handles the classification? using crowdsouring?
... we can use tags to classify
... some predefined classifications don't work well with some cases

<ericstephan> is this a real world example of terminology reuse?

Giancarlo_Guizzardi: is it useful to differentiate the feedback on quality?

<RiccardoAlbertoni> +1 to Giancarlo_Guizzardi yes we have to distingush at least quality feedback

phila: if it is useful, yes.
... do we have two types of Feedback or more types?
... we need need a diagram with classes and properties of both vocabs in one

BernadetteLoscio: we should keep the QualityFeedback to glue with DQV and investigate other types of feedback

<ericstephan> +1 phila!

BernadetteLoscio: there is a difference in the way the two vocabs use the dcat:dataset and dcat:distribution

phila: I can create the combined diagram now

<SumitPurohit> i am here

BernadetteLoscio: we can talk about the Citation part

ericstephan: in the next release of DUV, we should mention things about quality

Giancarlo_Guizzardi: we should go ahead with what Bernadette proposed but we can give a thought to other types
... same applies for motivation

<RiccardoAlbertoni> +1 to Giancarlo_Guizzardi and BernadetteLoscio proposal, do not enumerate all the classes

<ericstephan> +1 Giancarlo_Guizzardi

<SumitPurohit> yes

<yaso> ==15 min break==

<ericstephan> My stomach just growled

<ericstephan> at the mention of brownies :-)

<phila> OK, combined model is at https://www.w3.org/2013/dwbp/wiki/Dataset_usage_vocab_workspace#Sao_Paulo_Meeting

<ericstephan> thank you nandana

<RiccardoAlbertoni> good weekend nandana ..

<ericstephan> I like your combined model phila!

<yaso> thanks nandana

<yaso> next session: BernadetteLoscio will chair and I'll scribe

<ericstephan> Berna and Sumit - I really think we need to refine duv:Usage a bit

<BernadetteLoscio> Thanks Nandana!

<ericstephan> I don't think we need prov:Activity, prov:Association to do what we need to do.

<ericstephan> do we have time?

<BernadetteLoscio> :)

<ericstephan> Giancarlo_Guizzardi if we removed prov:Activity and prov:Association between duv:Usage and prov:Agent would that make sense to you?

<yaso> chair: BernadetteLoscio

<yaso> scribe: yaso

<scribe> scribeNicK: yaso

BernadetteLoscio: ericstephan can you describe your proposal?

ericstephan: I'm thinking if it in someways we created our own duv usage
... i'm curious if anybody has reactions

Giancarlo_Guizzardi: it makes sense to me to have a usage that represents itself
... if we specialize, if we have our own DUV usage class
... and I think that it would be all we need

BernadetteLoscio: phil made a proposal to connect the 2 vocabularies

phila: it's not complete

<phila> https://www.w3.org/2013/dwbp/wiki/Dataset_usage_vocab_workspace#Sao_Paulo_Meeting

<ericstephan> https://www.w3.org/2013/dwbp/wiki/Dataset_usage_vocab_workspace#Sao_Paulo_Meeting

BernadetteLoscio: the quality annotation is missing

RiccardoAlbertoni: actually its like if we had it
... that's why I'm saying that in terms of incompatibility i don't see many issues
... we selected some different (?) but this is not an issue, I think

<phila> Updated diagram

<phila> OK, try this https://www.w3.org/2013/dwbp/wiki/File:Bothvocabs.png

<ericstephan> Very nice phila !

<RiccardoAlbertoni> Ok, thanks phil the exercise is actually useful :)

SumitPurohit: before we explain more about it, has anyone any comment about it?

BernadetteLoscio: we are reusing something, I'm not sure if CITO is a standard, if we should reuse


BernadetteLoscio: because we can define our own class for citation

<ericstephan> http://sempublishing.sourceforge.net

ericstephan: I don't know if people are familiar with
... I think CITO is sufficient

<SumitPurohit> +q

<Giancarlo_Guizzardi> CITO has an object property called "cites as data source"

SumitPurohit: in my head, the 1st question is:can we loose something using CITO?

phila: if we know this is widely used and the field is very good

Giancarlo_Guizzardi: i did know CiOT and there's interesting things here
... i'm sure that you already mentioned this, bit the relation between
... citation act is a per-formative act, which is quite general
... maybe this is a way to represent by feedback
... and more suitable to do what we what that it does

<phila> CrossRef search based on a DOI

<ericstephan> +1 I agree

<SumitPurohit> +q

<BernadetteLoscio> thanks Bart!

<ericstephan> +1 Sumit to not going too generalized

<phila> :thisCitation

<phila> a duv:Citation;

<phila> cito:hasCitingEntity :dataset-03312004;

<phila> cito:hasCitedEntity :paperA;

<phila> .

<SumitPurohit> +q

<ericstephan> http://www.essepuntato.it/lode/http://purl.org/spar/cito#objectproperties

<SumitPurohit> +1

<SumitPurohit> +1 to phil

<Caroline> +1 to phil

<SumitPurohit> +q


<ericstephan> http://www.essepuntato.it/lode/http://purl.org/spar/cito#objectproperties

<phila> ACTION: sumit to investigate the relationship between DQV and DUV wrt citations that can be considered as a quality annotation [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action06]

<trackbot> Created ACTION-210 - Investigate the relationship between dqv and duv wrt citations that can be considered as a quality annotation [on Sumit Purohit - due 2015-10-02].

<ericstephan> +1

<nandana> +1 to RiccardoAlbertoni. It seems citations are a bit similar to incoming and outgoing links in data. At least we measure those a lot for quality.

HCLS Dataset Description

<phila> http://www.w3.org/2001/sw/hcls/notes/hcls-dataset/

<ericstephan> it uses cito

<phila> Yes, it uses CiTO, PAV, DCAT etc.

<phila> And it covers versioning etc.

Wrapping Up

<phila> ericstephan: We need to incorporate all the comments and get a new version out of the door as soon as we can.

<phila> BernadetteLoscio: We're planning a new draft of the docs in early November

<ericstephan> perhaps even before tpac?

<phila> BernadetteLoscio: Thanks everyone!

<newtonca_> may we close action-204 and action-206 ?

<ericstephan> Great job everyone and safe travels!

thanks phila and phila for coming!

<phila> deirdrelee: Thanks everyone for joining in over the last two days. We have got through a lot, closed a lot of issues etc.

<phila> ... Not it's the home stretch

<Giancarlo_Guizzardi> Thanks, Eric. This was very productive and fun

<phila> ... need to get the next drafts out. CR etc.

and thanks for those who attended remotely :-D

<ericstephan> Thank you Giancarlo_Guizzardi wow loved your insights!

<RiccardoAlbertoni> Thank guys. it is always very interesting discuss with the group :) Enjoy drinks and sao paolo

<annette_g> thanks, guys!

<Giancarlo_Guizzardi> Great! I appreciate that. Looking forward to interact more with you guys in the future

<nandana> Thanks and bye all !!

<Giancarlo_Guizzardi> Bye!

<annette_g> bye!

<RiccardoAlbertoni> bye!

<newtonca_> close action-204

<trackbot> Closed action-204.

<newtonca_> close action-206

<trackbot> Closed action-206.

Summary of Action Items

[NEW] ACTION: antoine to contact OA WG to see whether they would consider adding DQV motivation - due 16 October [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action04]
[NEW] ACTION: Bernadette to tabulate requirements against the BPs that address them [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action01]
[NEW] ACTION: calegari to incorporate the glossary in the BP document [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action02]
[NEW] ACTION: ericstephan to contact OA WG to see whether they would consider adding DUV motivation - due 16 October [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action05]
[NEW] ACTION: Gisele to modify the Data Enrichment Best Practice to cover data as well as metadata [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action03]
[NEW] ACTION: sumit to investigate the relationship between DQV and DUV wrt citations that can be considered as a quality annotation [recorded in http://www.w3.org/2015/09/25-dwbp-minutes.html#action06]
[End of minutes]