Chatlog 2010-04-29 Data Catalog Vocabulary
From W3C eGovernment Wiki
Please justify/explain all edits to this page, in your "edit summary" text.
<sandro> Guest: Bobbin Teegarden, <sandro> Guest: William (ww) Waites <sandro> Guest: Paul (PaulZH) Hermans <sandro> Present: Brand_Niemann, David_James, Ed_Summers, Fadi_Maali, George_Thomas, Jon_Phipps, Li_Ding, Luigi_Montanez, Martín_Álvarez, Michael_Hausenblas, Richard_Cyganiak, Vassilios Peristeras 15:01:16 <RRSAgent> RRSAgent has joined #egov 15:01:16 <RRSAgent> logging to http://www.w3.org/2010/04/29-egov-irc 15:01:18 <trackbot> RRSAgent, make logs public 15:01:20 <trackbot> Zakim, this will be EGOV 15:01:20 <Zakim> ok, trackbot, I see T&S_EGOV(LD TECH)10:00AM already started 15:01:21 <trackbot> Meeting: eGovernment Interest Group Teleconference 15:01:21 <trackbot> Date: 29 April 2010 15:01:30 <cygri> nice. thanks mhausenblas! 15:01:54 <cygri> agenda: http://www.w3.org/egov/wiki/Data_Catalog_Vocabulary/2010-04-29 15:02:06 <edsu> Zakim, LC is edsu 15:02:06 <Zakim> +edsu; got it 15:02:10 <mhausenblas> RRSAgent, draft minutes 15:02:10 <RRSAgent> I have made the request to generate http://www.w3.org/2010/04/29-egov-minutes.html mhausenblas 15:02:12 <Zakim> +??P7 15:02:25 <cygri> Chair: cygri 15:02:31 <cygri> Scribe: edsu 15:02:52 <cygri> zakim, who is here? 15:02:52 <Zakim> On the phone I see ??P3, +1.202.955.aaaa, martin, George, +03539149aabb (muted), edsu, ??P7 15:02:54 <Zakim> On IRC I see RRSAgent, Zakim, LuigiMontanez, PaulZH, george, jonphipps, martin, fadi, cygri, n8fr8__, mhausenblas, hughb, ww, edsu, sandro, trackbot 15:03:10 <Zakim> +jonphipps 15:03:29 <lindstream> lindstream has joined #egov 15:03:32 <Zakim> +??P8 15:03:57 <n8fr8__> n8fr8__ has left #egov 15:04:03 <Zakim> - +03539149aabb 15:04:09 <vassilios> vassilios has joined #egov 15:04:12 <DavidJames> DavidJames has joined #egov 15:04:46 <edsu> that's dc area 15:04:54 <Zakim> + +1.425.378.aacc 15:05:23 <Zakim> + +03539149aadd 15:06:00 <edsu> ww: are you on the call? 15:06:09 <PaulZH> I'm also calling IP 15:06:56 <bobbin> bobbin has joined #egov 15:06:57 <cygri> zakim, who is talking? 15:07:08 <Zakim> cygri, listening for 10 seconds I heard sound from the following: ??P3 (9%), ??P8 (9%) 15:07:18 <fadi> zakim, aabb is fadi 15:07:18 <Zakim> sorry, fadi, I do not recognize a party named 'aabb' 15:07:29 <cygri> zakim, who is on the phone? 15:07:29 <Zakim> On the phone I see ??P3, +1.202.955.aaaa, martin, George, edsu, ??P7, jonphipps, ??P8, +1.425.378.aacc (muted), +03539149aadd (muted) 15:07:46 <cygri> zakim, aadd is fadi 15:07:46 <Zakim> +fadi; got it 15:08:09 <vassilios> vassilios also on IP phone 15:09:08 <cygri> zakim, aaaa is LuigiMontanez 15:09:08 <Zakim> +LuigiMontanez; got it 15:09:16 <cygri> zakim, who is here? 15:09:16 <Zakim> On the phone I see ??P3, LuigiMontanez, martin, George, edsu, ??P7, jonphipps, ??P8, +1.425.378.aacc (muted), fadi (muted) 15:09:18 <Zakim> On IRC I see bobbin, DavidJames, vassilios, lindstream, RRSAgent, Zakim, LuigiMontanez, PaulZH, george, jonphipps, martin, fadi, cygri, mhausenblas, hughb, ww, edsu, sandro, 15:09:20 <Zakim> ... trackbot 15:09:58 <edsu> Topic: Admin 15:10:21 <edsu> cygri: agenda http://lists.w3.org/Archives/Public/public-egov-ig/2010Apr/0084.html 15:10:51 <edsu> ... we'll have a presentation from martin of ctic, and hopefully one from li ding of rpi 15:11:23 <edsu> ... time permitting i'd like to talk about the scope and deliverables of the group 15:11:34 <edsu> ... and hopefully write those down on the wiki 15:11:39 <edsu> Topic: Demos 15:12:13 <cygri> wiki page: http://www.w3.org/egov/wiki/Data_Catalog_Vocabulary/Datos_de_Asturias 15:12:19 <edsu> martin: here at ctic we develop an open data solution in spain, a good proof of concept of the linked data solution 15:12:59 <edsu> ... here you can see some metadata based on the VoID and Dublin Core vocabularies 15:13:24 <edsu> ... you can see license and provenance information, and the available data formats 15:13:47 <edsu> ... the generation of the catalog is using sparql and xslt 15:14:29 <martin> http://www.w3.org/egov/wiki/Data_Catalog_Vocabulary/CTIC_Dataset_List 15:14:30 <edsu> ... we have a list of catalogs (a meta catalog) 15:15:16 <fadi> fadi has joined #egov 15:15:23 <edsu> ... similar to the deri dcat approach, title, description, license 15:15:28 <martin> http://datos.fundacionctic.org/sandbox/catalog/feeds/catalogs_en.rss 15:15:40 <martin> http://datos.fundacionctic.org/sandbox/catalog/map.html.en 15:16:06 <edsu> ... also have a rss feed of the catalogs, which is used to geenrate a map of the catalogs 15:17:18 <edsu> cygri: you have used void for the data catalog, are the datasets in the catalog all available as rdf? this is not the case for most catalogs (e.g. data.gov) 15:17:38 <edsu> ... do you think it makes sense to use void when the dataset is not rdf? 15:18:21 <edsu> martin: i think void is by definition only for rdf data sets ... so i think no? 15:18:52 <edsu> cygri: that's what void was made for originally, but it possibly could be applied in a broader scope ... but i take your answer as saying you'd only use it for rdf data sets? 15:19:04 <edsu> martin: i think void could be useful 15:19:23 <cygri> zakim, who is here 15:19:23 <Zakim> cygri, you need to end that query with '?' 15:19:29 <edsu> mapstraction.addOverlay('http://datos.fundacionctic.org/sandbox/catalog/feeds/catalogs_en.rss'); 15:19:35 <cygri> zakim, who is here? 15:19:35 <Zakim> On the phone I see ??P3, LuigiMontanez, martin, George, edsu, ??P7, jonphipps, ??P8, +1.425.378.aacc (muted), fadi (muted) 15:19:37 <Zakim> On IRC I see fadi, bobbin, DavidJames, vassilios, lindstream, RRSAgent, Zakim, LuigiMontanez, PaulZH, george, jonphipps, martin, cygri, mhausenblas, hughb, ww, edsu, sandro, 15:19:40 <Zakim> ... trackbot 15:20:34 <Zakim> + +1.202.564.aaee 15:21:27 <edsu> cygri: do we have a new attendee from dc? 15:21:48 <edsu> cygri: ok, it's brandt from epa in washington 15:21:54 <cygri> zakim, aaee is brand 15:21:54 <Zakim> +brand; got it 15:23:04 <edsu> bobbin: i'm on the call too 15:23:15 <cygri> zakim, aacc is bobbin 15:23:15 <Zakim> +bobbin; got it 15:23:56 <cygri> zakim, who is here? 15:23:56 <Zakim> On the phone I see ??P3, LuigiMontanez, martin, George, edsu, ??P7, jonphipps, ??P8, bobbin, fadi, brand 15:23:59 <Zakim> On IRC I see fadi, bobbin, DavidJames, vassilios, lindstream, RRSAgent, Zakim, LuigiMontanez, PaulZH, george, jonphipps, martin, cygri, mhausenblas, hughb, ww, edsu, sandro, 15:24:01 <Zakim> ... trackbot 15:24:25 <DavidJames> LuigiMontanez and I are sharing a conference phone 15:24:43 <cygri> zakim, davidjames is with luigimontane 15:24:43 <Zakim> +davidjames; got it 15:24:50 <cygri> zakim, davidjames is with luigimontanez 15:24:50 <Zakim> davidjames was already listed in LuigiMontanez, cygri 15:24:51 <bobbin> Bobbin Teegarden, semantic consultant in Seattle 15:25:14 <edsu> cygri: seems like Li isn't going to make the call 15:25:21 <edsu> Topic: Deliverables 15:26:37 <edsu> cygri: there wasn't a clear consensus at the end of the last call, about defining an abstract vocabulary about the data you find in catalogs, and then we could have several different documents that explain how you could use these terms in a concrete syntax: rdf/xml, rdfa, atom, etc 15:26:54 <edsu> ... a couple of points were raised on the mailing list 15:27:13 <edsu> ... erik suggested we should be clear about the types of serices we want to enable 15:27:27 <edsu> ... so that's where i would like to pick up 15:27:40 <DavidJames> we're getting a lot of echo and delay on the line 15:27:41 <LI_DING> LI_DING has joined #egov 15:28:20 <edsu> cygri: i sent an email http://lists.w3.org/Archives/Public/public-egov-ig/2010Apr/0083.html 15:28:50 <edsu> ... one deliverable might be use cases and requirements 15:29:20 <edsu> ... a second thing could be an rdf vocabulary and reference that lists the terms, which required/optional, examples 15:29:56 <edsu> ... it would make sense to use some sort of formalism, uml, or perhaps rdf schema, since it will give us one possible way of deploying dcat data 15:30:55 <edsu> ... in the email the choice of concrete format maybe we should not decide on this yet ... if we have some clear requirements for using syndication formats for example ; or if we can do everything with rdfa -- it's hard to say before we know what requirements we have 15:31:23 <edsu> ... do you think we could delay the decision about the concrete format until we do use cases requuirements? 15:31:23 <Zakim> + +1.518.495.aaff 15:31:34 <Zakim> - +1.518.495.aaff 15:31:50 <PaulZH> I do agree to postpone the decision 15:32:38 <Zakim> + +1.518.495.aagg 15:32:47 <edsu> edsu: i agree in principle, but it will take more time 15:32:52 <LI_DING> hi 15:32:55 <LI_DING> li is joining 15:33:25 <edsu> Zakim, +1.518.495.aagg is LI_DING 15:33:25 <Zakim> +LI_DING; got it 15:34:04 <edsu> LI_DING: we're reviewing cygri's proposal at http://lists.w3.org/Archives/Public/public-egov-ig/2010Apr/0083.html 15:34:55 <edsu> cygri: i think by writing down use cases and requirements we'll get a better understanding about technical choices 15:35:30 <edsu> LI_DING: what is the current status, discussion the requirements? 15:36:27 <edsu> cygri: yes, basically we're discussing whether we should abstract out of the concrete syntax, and write use cases and requirements 15:37:22 <edsu> ... there doesn't seem to be a lot of disagreement or agreement with this 15:37:55 <edsu> LI_DING: i want to hear more voices first i think 15:38:11 <cygri> zakim, who is on the phone 15:38:11 <Zakim> I don't understand 'who is on the phone', cygri 15:38:23 <cygri> zakim, who is on the phone? 15:38:23 <Zakim> On the phone I see ??P3, LuigiMontanez, martin, George, edsu, ??P7, jonphipps (muted), ??P8, bobbin, fadi (muted), brand, LI_DING 15:38:25 <Zakim> LuigiMontanez has LuigiMontanez, davidjames 15:38:58 <DavidJames> Regarding http://lists.w3.org/Archives/Public/public-egov-ig/2010Apr/0083.html ... I like the list of requirements. 15:39:49 <DavidJames> The deliverables also look good to me. 15:39:55 <edsu> LI_DING: i'd need more time to review the proposal, but i think we want to keep things as minimal as possible to get something out there quick, also i think we should keep use cases simple, how much do we want to get? 15:40:19 <edsu> ... we should try to produce something actionable 15:41:37 <edsu> cygri: i don't hear any strong objection to moving forwards with the deliverables i outlined in the email ; for now we should continue working with this assumption 15:41:54 <LI_DING> one thing could be a challenge 15:42:03 <DavidJames> Can we ask everyone to vote (or something similar)? 15:42:34 <edsu> LI_DING: maybe we can have a challenge a deliverable: how many gov't data sets can you find? 15:43:16 <edsu> ... lets grow the number of government data sets, more people using the vocabulary 15:43:45 <edsu> cygri: but what does this mean for the format we want to design? 15:44:10 <vassilios> i think it is not a matter of number of datasets but number of catalogues 15:45:07 <Zakim> -brand 15:45:22 <edsu> LI_DING: it's a parallel effort: one side we grow the datasets, and on the other we develop the terminology, vocabulary that can be used 15:46:16 <edsu> cygri: DavidJames you asked whether we should do a vote, i'm not sure what point this would serve 15:46:33 <edsu> ... since there doesn't seem to be any strong disagreement 15:46:43 <cygri> ack DavidJames 15:46:48 <edsu> DavidJames: i think it would be nice to hear the agreement 15:47:22 <edsu> cygri: i think this is something i'd like to do on the next call, rather than rushing it 15:47:35 <edsu> ... to make sure there is enough time 15:47:49 <edsu> Topic: Use Cases 15:48:56 <edsu> cygri: we want to have use case or stories for what people want to be able to do: for example i want to be able to get a list of datasets from data.gov that are not supported by the current data.gov user interface 15:49:22 <edsu> ... i'd be interested to hear about what use cases people on the call have 15:49:32 <edsu> ... for standard formats for data catalogs 15:49:42 <Zakim> -LI_DING 15:50:41 <edsu> martin: i'd like to be able to query multiple catalogs, to be able to query other geographic regions 15:51:16 <vassilios> cross-catalogue querying 15:51:25 <edsu> cygri: so you want to be able to federate national catalogs into say a european union catalog? 15:51:47 <edsu> s/martin/vassilios/ 15:52:00 <edsu> vassilios: yes 15:52:21 <george> (fed/state/local) gov agency independently published catalog using (some set of standard) vocab for dynamic aggregation of catalogs 15:52:55 <george> would like to see bottom up (as per Li - in use) drive basic (extensible) rdfs vocab convergence for w3c blessing 15:53:00 <edsu> DavidJames: for the nationaldatacatalog we'd like to consume data from other data catalogs, we'd like to have one uniform way of getting at catalog data 15:53:33 <edsu> ... for municipalities and state governments we'd like to do a similar thing, enabling federation across many levels 15:54:38 <edsu> george: i think i share the interest that david and luigi have ; would be great to have some sort of vocabulary at the w3c that reflects what's in use as much as possible 15:54:42 <edsu> cygri: agreed 15:55:47 <edsu> martin: i think federation is soemthing i would like to do ; i agree with vassilios with his approach 15:55:56 <LI_DING> I was dropped from the call 15:56:46 <edsu> cygri: it might be a good next step to draft this use case as a little document, to sketch how it could look in practice as a story 15:56:47 <vassilios> i could help on this 15:57:14 <edsu> ... there are a couple of other use cases on the mailing list, maybe we can get those into written form too 15:57:27 <vassilios> OK 15:57:55 <vassilios> and i thin kit is more or less the same whether we speak about EU or USA 15:58:03 <edsu> cygri: would be great if you could start it, and we could see if other use cases are covered 15:58:33 <edsu> ACTION: vassilios to document a federation use case for dcat 15:58:33 <trackbot> Could not create new action (unparseable data in server response: No child element named 'id') - please contact sysreq with the details of what happened. 15:58:48 <george> +1 15:58:48 <LuigiMontanez> yes, I have access 15:58:50 <LuigiMontanez> +1 15:58:52 <DavidJames> +1 15:58:56 <edsu> +1 15:58:56 <martin> +1 15:58:58 <fadi> +1 15:59:03 <edsu> cygri: wiki is ok? 15:59:17 <cygri> http://www.w3.org/egov/wiki/Data_Catalog_Vocabulary/Use_Cases_and_Requirements 15:59:36 <edsu> cygri: great, there is a page where i've copy pasted stuff from the email, i propose we add subsections there, use this as a live document 15:59:53 <edsu> cygri: so that's it for this week, thank you all for joining in 15:59:56 <DavidJames> thanks 16:00:00 <george> thank you cygri 16:00:01 <cygri> thanks all! 16:00:06 <LuigiMontanez> thank you cygri, edsu for transcribing 16:00:06 <vassilios> thank you! 16:00:13 <Zakim> -George 16:00:17 <Zakim> -??P3 16:00:18 <george> thanks edsu 16:00:21 <Zakim> -LuigiMontanez 16:00:23 <Zakim> -fadi 16:00:25 <Zakim> -jonphipps 16:00:25 <edsu> np :-) 16:00:29 <Zakim> -edsu 16:00:31 <Zakim> -??P8 16:00:37 <Zakim> -martin 16:00:39 <edsu> RRSAgent, generate minutes 16:00:39 <RRSAgent> I have made the request to generate http://www.w3.org/2010/04/29-egov-minutes.html edsu 16:00:51 <Zakim> -bobbin 16:01:05 <cygri> ok, chairing a teleconference from a conference hallway is a bad idea. i won't try that again :-) 16:01:19 <edsu> cygri: heheh, any news from www2010? 16:01:21 <george> wish i was there! 16:01:29 <LuigiMontanez> LuigiMontanez has joined #egov 16:01:36 <bobbin> zakim? 16:02:08 <cygri> edsu, yes, many interesting things going on. great conference so far 16:02:10 <edsu> shoot, the minutes don't look to good 16:02:17 <edsu> http://www.w3.org/2010/04/29-egov-minutes.html 16:02:31 <cygri> bobbin, what do you want zakim to do? 16:02:51 <bobbin> +1 16:03:19 <edsu> cygri: it does look good 16:03:26 <edsu> (the conference) 16:03:40 <edsu> cygri: have you and Li had a chance to chat about stuff informally? 16:03:40 <cygri> edsu, the minutes look good to me? 16:03:51 <edsu> oh, they do look better now when i reloaded 16:04:02 <cygri> edsu, yes we had dinner yesterday and are planning a few things 16:04:08 <edsu> cygri: nice 16:04:26 <edsu> cygri: i think having you and Li on the same page will help a great deal 16:04:30 <Zakim> -??P7 16:04:31 <Zakim> T&S_EGOV(LD TECH)10:00AM has ended 16:04:32 <Zakim> Attendees were +1.202.955.aaaa, martin, George, +03539149aabb, edsu, jonphipps, +1.425.378.aacc, +03539149aadd, fadi, LuigiMontanez, +1.202.564.aaee, brand, bobbin, davidjames, 16:04:35 <Zakim> ... +1.518.495.aaff, LI_DING 16:05:19 <DavidJames> DavidJames has left #egov 16:06:27 <cygri> edsu, agreed 16:07:03 <cygri> will go now and see if they left any lunch for me 16:07:10 <cygri> thanks for scribing and the comments ed 16:18:33 <lindstream> lindstream has left #egov 17:16:50 <cygri> cygri has joined #egov 17:40:23 <ww> edsu: sorry wasn't able to make it today 17:41:05 <ww> would have liked to bring up the spatial/temporal ambiguity of dcat:granularity though 17:50:49 <DavidJames> DavidJames has joined #egov 17:56:42 <ww> cygri: do you think it would be reasonable to distinguish between spatial and temporal granularity/resolution in dcat? 17:58:11 <ww> (nevermind datasets that have different granularity on their different axes, which happens with some envionmental satellite datasets for example) 17:58:12 <cygri> ww, i'll have to talk to fadi to remember why we didn't distinguish them. can i get back to you about this next week? 17:58:27 <ww> sure no hurry 18:01:24 <Zakim> Zakim has left #egov 18:04:36 <ww> any thoughts on rdf:graph, rdf:Graph? 18:09:50 <cygri> ww, what are those? 18:10:21 <ww> well, they don't exist, which I think is problematic 18:10:51 <ww> using the talis changeset, vocabulary, for example, changes are expressed as additions and removals of reified triples 18:11:10 <ww> i need a way to express the graph that those triples are supposed to live in 18:11:19 <ww> (not strictly related to egov as such, of couse) 18:11:34 <ww> we have rdf:Statement, rdf:subject, rdf:predicate, rdf:object 18:11:40 <ww> but no rdf:graph 18:12:03 <ww> I suppose because I believe that named graphs were added after the rdf core vocabulary was etched in stone 18:13:19 <ww> for my immediate purposes I've invented a namespace for this, but... 19:33:35 <DavidJames> DavidJames has joined #egov 19:44:52 <cygri> cygri has joined #egov 20:54:51 <cygri> cygri has joined #egov 21:00:21 <cygri> cygri has joined #egov # SPECIAL MARKER FOR CHATSYNC. DO NOT EDIT THIS LINE OR BELOW. SRCLINESUSED=00000253