– DRAFT –
DCAT team 2018-08-16

<DaveBrowning> rrsagent: make logs public

review agenda

scribe+ arminhaller

agenda approved

approve minutes https://‌www.w3.org/‌2018/‌08/‌09-dxwgdcat-minutes

<alejandra> +1

<DaveBrowning> +1

+1

<SimonCox> missed some regrets:

<roba> +1

<SimonCox> andrea for last time

<SimonCox> +1

approved minutes

Outstanding actions https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌open

<SimonCox> ?

<Makx> yes close #84

Revision of milestone https://‌github.com/‌w3c/‌dxwg/‌issues?q=is%3Aopen+is%3Aissue+milestone%3A%22DCAT+Second+Public+Working+Draft%22

<DaveBrowning> https://‌github.com/‌w3c/‌dxwg/‌milestone/‌6

DaveBrowning: Issues around milestones that are marked as 0% complete, but are closed
… suggestion to link the wiki page to each milestone, where the justification is for why the milestone was closed

<SimonCox> link to wiki, else to a issue group like https://‌github.com/‌w3c/‌dxwg/‌issues?utf8=%E2%9C%93&q=is%3Aissue+label%3Aversion+

DaveBrowning: wanted to hear opinions on that

<SimonCox> I closed all the topic-oriented milestones. Perhaps I got the github flow wrong ...

alejandra: two criteria for creating a milestone. one per topic, grouping issues, for example, for versioning. We changed recently, milestones now point to products, i.e. the next public working draft.

SimonCox: Agree with alejandra, milestones were being used for different purposes initially.
… I closed them, but I should have probably emptied them first. So that they don't have open issues.

DaveBrowning: I am quite happy to add those wiki page links
… please hold off your tidying

Action: SimonCox to clean up the milestons on github after Dave is done with the editing changes

<trackbot> Sorry, but no Tracker is associated with this channel.

Public comments list: https://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-comments/

alejandra: There were more comments on the public mailing list. Simon responded to Melanie.

SimonCox: Not sure if I properly addressed the concerns she had.

Profile description

<alejandra> sub-topic: Revisit discussion about profile description being broader and independent of DCAT

roba: In the plenary there was a discussion around the requirements for the profile description ontology
… when we brought this to this group the discussion was that the profiles should have their own document, as they apply not only to DCAT but more generally
… but I can't find the documentation on this decision. I wanted to get the group to confirm this decision
… since this discussion we have also generalised a dataset to be a subclass of dcat:resource

<SimonCox> Can you make a concrete proposal roba ?

alejandra: I have a vague recollection of this discussion

arminhaller: I can recall that conversation too.

alejandra: a profile could be catalogued
… a profile could be a resource

arminhaller: +1 on profile becoming a dcat:resource

<roba> PROPOSED: The DCAT sub-group agrees that the description of profiles is relevant to meeting some requirements for DCAT but since it is broader conceptually than the DCAT scope should be addressed in the profile guidance deliverables

<SimonCox> +1

+1

<DaveBrowning> +1

<roba> +1

<alejandra> +1

<Makx> +1

<alejandra> sub-topic: Discussion on whether DCAT should include a module for cataloguing profile descriptions, as per the DCAT-AP use case https://‌github.com/‌w3c/‌dxwg/‌issues/‌238

<roba> https://‌github.com/‌w3c/‌dxwg/‌blob/‌gh-pages/‌profiledesc/‌profiledesc_dcat_alignment.ttl

<SimonCox> s/since is broader/since it is broader/

arminhaller: @base <https://‌w3c.github.io/‌dxwg/‌profiledes> misses the slash at the end

<alejandra> is cataloguing profiles in scope?

arminhaller: or actually # and a 'c' at the end

arminhaller: +1 on first module

<SimonCox> is 'Profile rdfs:subClassOf dct:Standard` ? (I'm thinking about whether/how to include it in the Figure 1 diagram ...)

alejandra: other modules align dcat with something else

roba: yes, agree with SimonCox

<Makx> Making profile a subclass of dct:Standard makes sense

SimonCox: The current summary in Figure 1 is incomplete

roba: Should be up to this group and then bring it back to the plenary
… if the proposal is to explore to further develop the concept of a profile as a catalogue resource, we can do it here

SimonCox: not sure if I am happy to make the decision now
… the scope of DCAT could be a second diagram, that is informative. But I am not sure yet.

roba: if there is a home for the action, I am happy to do it. Maybe assign an action to me.

alejandra: I am in favour of having the ontology, I am not sure it should be in the spec
… same as with services, we say that we can catalogue them, but we don't make recommendations

roba: happy to do the same for profiles as for services

<Makx> bundling all of it together makes the spec harder to use.

roba: the way you catalogue services can be used for profiles

SimonCox: services are already in there, so what you propose for profiles makes sense

Resolved: DCAT determines that the description of profiles is relevant to meeting some requirements for DCAT but since is broader conceptually than the DCAT scope should be addressed in the profile guidance deliverables

<roba> PROPOSED: treat profiles at the same level of detail and using the same mechanism as services in the DCAT specification

<alejandra> +1

<DaveBrowning> +1

<roba> +1

<Makx> +1

+1

<SimonCox> +1

Resolved: treat profiles at the same level of detail and using the same mechanism as services in the DCAT specification

Resolved: treat profiles at the same level of detail and using the same mechanism as services in the DCAT specification

Data citation

<alejandra> https://‌github.com/‌w3c/‌dxwg/‌issues/‌61

<alejandra> wiki page: https://‌github.com/‌w3c/‌dxwg/‌wiki/‌Data-Citation

alejandra: I am preparing a pull request that included a creator and identifier for a dataset

<Makx> alejandra, there is already identifier for dataset in DCAT 2014: https://‌www.w3.org/‌TR/‌vocab-dcat/#Property:dataset_identifier

SimonCox: identifier is present. creator is not present. title is, publisher is, resourcetype is, publicationYear is

<alejandra> thanks Makx

SimonCox: the point here is to make all of the properties mandatory

<Makx> Making properties mandatory sounds like a profile

alejandra: yes, the proposal is to make them part of the profile description

+1 agree with Makx

<DaveBrowning> +1 to Makx

<SimonCox> e.g. the DataCite profile would make identifier, title, publisher, type, creator, issued all mandatory

<alejandra> correct

roba: It is perfectly reasonable that we say that this requirement can be met by a profile

SimonCox: we are not publishing profiles
… anything in this regard needs to be non-normative

<Makx> +1 to Simon: not publishing profiles

roba: Yes, we make the recommendation and potentially an example, but informative

DaveBrowning: Agree with the summary. There are quite a few requirements where we can say that they are met with a profile
… it may become a bit tedious in the document, if we say all the time that this is not met by DCAT, and you have to use a profile

alejandra: agree

<SimonCox> Currently we have this placeholder section: https://‌w3c.github.io/‌dxwg/‌dcat/#alignments

alejandra: it should not be part of the spec, but this is a nice example (i.e. the DataCite one) that we could use

<Zakim> SimonCox, you wanted to propose maybe a series of annexes in the DCAT rec?

<alejandra> SimonCox: datacite alignment is a combination of mapping and profile

alejandra: wondering whether they are normative
… maybe have some annexes

alejandra: normative can be anything that is in the charter

<alejandra> ?

<Makx> I would be reluctant to have lots of annexes. the DCAT spec should be a short as possible.

<alejandra> arminhaller: I can answer that question partly

<alejandra> ... a section can be normative even if it is not in the charter

<Makx> examples of profiles could be in a separate profile document

<alejandra> ... we made some arbitrary decision about what we included as non-normative

<SimonCox> +1 to armin's suggestion that we can decide what is normative ...

<alejandra> ... the process of whoever was willing to do the alignment

<alejandra> ... this is a really good example as profile

<alejandra> ... we can make a decision on how to include it

roba: the normative part is in the solution space
… it is the decision of the group what you make normative
… a decision how you meet you requirements

alejandra: i can take an action to continue on the DataCite example

<SimonCox> FYI potential taxonomy of dcat:Resources: https://‌github.com/‌w3c/‌dxwg/‌blob/‌gh-pages/‌dcat/‌UML/‌Resources.png

alejandra: DaveBrowning will chair next week