Government Linked Data Working Group Teleconference -- 15 Mar 2012

<trackbot> Date: 15 March 2012

<bhyland> Linked Data Glossary, see https://dvcs.w3.org/hg/gld/raw-file/default/glossary/index.html

<olyerickson> ??p21 is me

<luis_bermudez> Hi. I will connect via phone in 10 min

<olyerickson> zqakim, ??p21 is me.

<bhyland> scribe: DaveReynolds

<scribe> scribenick: DaveReynolds

<bhyland> Propose to accept minutes: http://www.w3.org/2011/gld/meeting/2012-03-08

<boris> +1

<olyerickson> +1 accept minutes

<gatemezi> +1 accept minutes

<fadmaa> +1

<DanG> +1

bhyland: requested people, dcat, org, data cube to first pass WD

<PhilA2> Terrific news! Well done to the folk who've worked hard on those

<PhilA2> Let me start with http://www.w3.org/QA/2012/03/interoperable_governments.html

PhilA update on ADMS and ISA vocabs

Update on the ADMS spec and ISA eGov core vocabs

PhilA2: ISA producing three core vocabs.
... People - soley a natural person (not businesses-as-people etc)
... business - to describe legal entities, subsets of organizations
... location - address information, intended to be INSPIRE compliant
... ISA person can be used as a subset of W3C people vocab
... ISA business not directly connected to org (though compatible?)
... Fourth vocab. ADMS. Asset Descriptions Metadata Schema.
... It is relevant for GLD. Government institutions are publishing assets like code lists, standards, vocabs. ADMS is for describing those (c.f. DCAT for describing datasets themselves).

<PhilA2> http://joinup.ec.europa.eu/

PhilA2: ADMS being implemented by joinup platform

<PhilA2> http://digitaliser.dk/

PhilA2: ADMS also being used by Danish portal
... W3C will also use AMDS for TR descriptions
... Now suggesting that GLD take on all four of the vocabularies.
... All of them match existing GLD charter.
... Are being asked to take result of EU review process as FPWD-ready material. But are, of course, free to then extend and improve before final publication.
... All four are intended to ready end of March for targetted FPWD publication in May.
... There is a 5th vocab, for description of open source software. Not being brought to GLD at this time (outside of charter)
... It is called RADion - Repository Assets Description

PhilA2 - very high level, DCAT could be seen as specialization of it (?)

<PhilA2> -> RADion details http://lists.w3.org/Archives/Public/public-gld-wg/2012Mar/0048.html

bhyland: a lot of the heavy lifting has been done on these vocabs, it is a funded effort, they are well documented. Encourages GLD folks to look them and voice opinions on them.

<olyerickson> I haven't see region-specific in things I've looked at

bhyland: especially - they should be checked for any region-specific aspects

<olyerickson> We've found core business vocab interesting, and will be looking at location at people as well

<olyerickson> ADMS is harder to figure out application scope. I get DCAT (obviously), I don't quite get use cases for ADMS

<olyerickson> Could PhilA elaborate on ADMS use case (or two)

sandro: asks about timeline guidance for reviewing the vocabs

<bhyland> Note: ADMS Working Group list, see https://joinup.ec.europa.eu/asset/adms/document/adms-working-group

PhilA2: targeting completion 1 week on Friday

<PhilA2> -> ADMS http://joinup.ec.europa.eu/asset/adms/home

PhilA2: Aiming that ADMS 1.0 becomes FPWD as is because it is already being implemented. So early identification of major issues would be important.
... Feels that FPWD should separate core model from implementation guidance, which are merged in ADMS 0.9/1.0

<bhyland> @PhilA, should we be looking at section 5.2 ADMS Core Elements in https://joinup.ec.europa.eu/asset/adms/wiki/adms-specification-v06

PhilA2: Some issues around text v. strings and flagging of machine translation

<DanG> This problem was solved in ISO/IEC 19773

<fadmaa> I am happy to write a comment to the ADMS

<DanG> Sorry 19773 is called Metadata Modules

PhilA2: In particular some of the implementation guide is Europe specific

<PhilA2> http://joinup.ec.europa.eu/blog/adms-version-09-now-available-joinup-consultation

<PhilA2> ADMS Namespace doc http://philarcher.org/adms/ADMS_v0.10.rdf

<PhilA2> Joinup http://joinup.ec.europa.eu/

PhilA2: [Clarifying ADMS use cases ...] Joinup is an example use. It provides "semantic assets" - includes vocabularies, code lists, standards. Provides discovery portal for public sector employees to find such assets.
... ADMS a vocab for describing these assets. Such an asset is different from a dataset.

<PhilA2> RADion UML http://philarcher.org/adms/RADion%20v0.05.png

<olyerickson> PhilA this actually is helpful...thanks...

<olyerickson> Yes

<George> ack

<Zakim> DaveReynolds, you wanted to ask about registry v. repository

<olyerickson> PhilA described some very futurey concepts IMHO...

<DanG> why is geography separated from SKOS concept scheme and concepts?

PhilA2: is intended to have a registry aspect, has some notion of status.
... but doesn't have a notion of endorsement by a particular community

DanG: asks why geographic coverage is separated from SKOS concept schemes. Typically geo areas are concept schemes and should be handled uniformly.

<olyerickson> I need DanG to give an example

PhilA2: previously had a notion of "domain". Could generalize to allow SKOS concepts to be used there.

<gatemezi> In Geonames, codes are skos:Concept

PhilA2: question to group, is it OK to publish as FPWD before sorting issues such as uniform use of concept schemes.

<gatemezi> link Geonames: http://www.geonames.org/ontology/documentation.html

DanG: preference, if group agrees then make changes first before FPWD publication.

<olyerickson> request that DanG elaborate on these issues in "ISSUES"

DanG: points out ISO 19773 has an approach to data types for multi-text, multi-string, multi-data. Type includes both datum and additional information (e.g. language, cultural identifiers).

<olyerickson> Thanks!

george: suggests capturing these on issues list

PhilA2: [Clarifying relationship of OSS vocab to DOAP] Interest is from UK and Spanish gov. If GLD feels it is relevant it *could* could to encompass that as well. In RADion there is a link to project which is doap:Project.

<olyerickson> ;)

bhyland: reluctant to to cover OSS area as well, seems beyond GLD charter.

PhilA2: agreed

<olyerickson> "GLD recommends stakeholder consider any and all vocabularies of potential interest, now and in the future..."

<olyerickson> Thanks PhilA!

PhilA2: appreciates GLD taking on the vocabs that it is adopting

DCAT

<bhyland> Action-37

<bhyland> We're all looking at: https://www.w3.org/2011/gld/track/products/2

<bhyland> (or should be ..)

fadmaa: issues 2,3 are easy and should be able to get consensus. Issue 9 is much the same as 7 and 8.

olyerickson: issue 2,3 trivial things left to close off.

bhyland: some issues are raised but not necessarily accepted as issues to be addressed (i.e. OPEN)

olyerickson: issue 2 resulted in dct:language being added (so solved?)

<bhyland> @George, perhaps you should move some of these to OPEN status and some to CLOSED if they've been addressed, with comment …

george: trying to get clarify on what the actual issue is so can be clear on what are solved, what are unclear, what are actionable.

olyerickson: issue 2 - added dct:language at both levels
... issue 3 - about time/date format, list agreed to prefer xsd:dateTime, but stakeholders wanted other forms, discussion still open

<scribe> ACTION: olyerickson to annotate and update issues 2, 3 on tracker [recorded in http://www.w3.org/2012/03/15-gld-minutes.html#action01]

<trackbot> Created ACTION-58 - Annotate and update issues 2, 3 on tracker [on John Erickson - due 2012-03-22].

<PhilA2> I'll make a proposal by mail to put ISSUE-3 to bed - I don't think it's a big issue and we know how to handle it

fadmaa: question is whether dct:Distribution needed (ans: yes), are subsclasses needed (may not be fully representative)
... want to differentiate between direct download links and web pages which eventually lead to download

<fadmaa> https://www.w3.org/2011/gld/track/issues/8

<PhilA2> George: Is that clearer, John?

<PhilA2> olyerickson: Yes, that helps

<PhilA2> George: So it looks like we can open actions 7, 8 and 9 and then come to a resolution?

<PhilA2> olyerickson: I think we need to create actions on those of us owning DCAT and come up with a specific alternative if that can be done

<PhilA2> ... it's a raesonable issue that Martin is raising

<PhilA2> ... we need to put something concrete before the GLD WG

<gatemezi> + 1 to Fadi point

<PhilA2> fadmaa: So we should put specific proposals on the ML?

<martin> +1

<PhilA2> George: That sounds good - it can go on next weeks agenda

<bhyland> +1

<PhilA2> scribe: PhilA2

<boris> +1 to mark issues as open

<scribe> ACTION: Fadi Maali to address issues 7 - 9 on the mailing list [recorded in http://www.w3.org/2012/03/15-gld-minutes.html#action02]

<trackbot> Created ACTION-59 - Maali to address issues 7 - 9 on the mailing list [on Fadi Maali - due 2012-03-22].

People

<George> PhilA2: patronymic name

<George> ... include Titles, Prefixes, etc.?

<George> ... no rules - but pushed for patronymic names

<George> ... and 'alternative' name (which isn't foaf:nick)

<olyerickson> +1 to formal alternativeName... alternative right now is skos:prefLabel, skos:altLabel

George: Let's look at Best Prac for vocabulary selection
... I noticed some mention of accidental subsumption
... does anyone else think this is a useful addition to the doc?
... If my hospital subject were to use a vCard address, it would be subsumed as a type vCard object
... this presrnts itself as a problem to possible users
... LD COmmunity is not very oriented towards this effect

<DaveReynolds> scribe: DaveReynolds

PhilA2: frequently find want to reuse a property but domain/range constraints on it cause unintended consequences
... some LD users work round that by separating inferences from direct assertions, pragmatics

<olyerickson> Linked Data is not about creating a perfect organization of knowledge

boris: worth mentioning in Best Practice vocab section

<bhyland> Sounds like we need a good glossary :-)

george: also some need for more clarity of description on "products" in the procurement section, e.g. Linked Data Server

<boris> yes

<bhyland> nice segway

<Mike_Pendleton> agreed George

LD Glossary see https://dvcs.w3.org/hg/gld/raw-file/default/glossary/index.html

bhyland: would like Best Practices etc to link to glossary. New entries, improved descriptions welcome.

<olyerickson> PhilA mentions LD Patterns WG

PhilA2: asks if group is aware of Linked Data platform, now "patterns", working group

<Zakim> PhilA, you wanted to talk about a linked data server

PhilA2: relevant to glossary definition of e.g. Linked Data Server

<boris> http://www.w3.org/2011/gld/wiki/Meetings:Telecon20120315

<PhilA2> -> Using Open data http://www.w3.org/2012/06/pmod/

PhilA2: workshop in June in Brussels - about *using* open data
... special focus on policy modelling

bhyland: asks US members to check travel possibilities (for TPAC?)
... to check of a GLD f2f is possible during TPAC

<gatemezi> thanks!

<PhilA2> regrets for next week

- DRAFT -

Government Linked Data Working Group Teleconference

15 Mar 2012

Attendees

Contents

PhilA update on ADMS and ISA vocabs

Update on the ADMS spec and ISA eGov core vocabs

DCAT

People

LD Glossary see https://dvcs.w3.org/hg/gld/raw-file/default/glossary/index.html

Summary of Action Items

Scribe.perl diagnostic output