W3C

- DRAFT -

eGovernment Interest Group Teleconference

31 Mar 2010

See also: IRC log

Attendees

Present
George, Chris, +0203334aaaa, Sandro, +03498429aabb, edsu, cygri, josema, Owen?, John?, Vassilios
Regrets
kevin, rachel, cory
Chair
josema
Scribe
edsu, Ed Summers

Contents


<josema> trackbot, start telcon

<trackbot> Date: 31 March 2010

<cygri> hi folks. is the call *now*? i can't dial in, it tells me the conference is restricted

<sandro> yeah.

<Chris> saem

<sandro> it's broken.

<Chris> *same

<John> Me too...

<Chris> bugger...

<josema> I don't even get the error message

<Chris> I can see the e-Gov IG note now - "Best Practices in using Technology.... Don't!" lol

<josema> zakim sees George but he's not on IRC, so we cannot ask him :(

<sandro> I wish.

<sandro> I'm trying to reach someone.

<josema> ok, thx, sandro

<Chris> I think I'm in...

<Chris> accepted the code

is the meeting over?

<sandro> No, the meeting is supposed to start now.

<josema> Chris, did you hear a beep?

<Chris> yup

<josema> hmmm....

<Chris> dead air atm - cause I think I'm alone - no - someone else is on - very faint

<sandro> Chris and I are on....

<sandro> seems to be working now.

<cygri> yup i'm in

<josema> scribe: edsu

Agenda adjustments and next meetings

josema: next meeting April 14th, need a scribe

Review of open actions

<josema> http://www.w3.org/egov/IG/track/actions/open

<sandro> I heard some voices in the distance there.

Projects: status report from coordinators

<josema> http://www.w3.org/egov/wiki/Projects

josema: george can you report a summary on your project?

george: there's been some discussion about conventions of use for some of the standards, such as dcat, void, etc
... and trying to evangelize demonstrations that people already have

josema: is there any expected deliverable?

george: we haven't converged on that at this point

josema: so what you have decided to is to act as an aggregator of different initiatives?

george: that's kind of the default yes
... a reflection of sandro's comment about energy and availability of people involved

josema: is there a sense of doing technical work, as well as a case that could be made for decision makers?

george: yes

josema: there is a need to make a value-proposition to poplicy makers

george: i think there is a good opportunity for that

josema: best practices for using web technologies, chris?

Chris: we tried focus groups for publishing pdf files, and using mobile technology ; got some response about pdf
... we'll maybe get something written about redaction
... also some conversations with adobe
... trying to identify where the interest is

josema: there is some concern about the participation in the groups, which is when we decided to let other organizers go on their own for a bit

<OwenAmbur> Lost my Skype connection

josema: hopefully focusing on very specific topics will help

Chris: yeah, we were thinking of putting out short how-tos
... and go from there

josema: we should clearer ideas about what we should have produce, given the charter expiring in october
... don't be afraid of publishing something
... a working draft can be almost anything ; we shouldn't be afraid of putting drafts out there ; and also people should feel free to use the wiki
... people seem afraid sometimes
... daniel isn't on the call today for best practices

Chris: Brian updated the social media project homepage http://www.w3.org/egov/wiki/SocMed

What's going on / coming up

josema: sandro could you tell us about fose? http://www.fose.com/Events/FOSE-2010/Home.aspx

sandro: i was on a panel replacing john, with rachel and daniel and kevi
... i'll paste the link for the slides, and the video into irc
... there wasn't a lot of time for questions, and it seemed like some people were excited about it

<sandro> http://civsourceonline.com/2010/03/24/open-data-movement-asks-if-you-open-it-who-will-come/

sandro: might get some people in the IG as a result

josema: any other news from people on the call?

Chris: i've got a paper accepted to talk about the Metadata Conference in Canberra

<OwenAmbur> The Industry Advisory Council (IAC) will be conducting a study of best practices with respect to the management of records created in social networking services

John: i've got a paper w/ Jeni Tennison to talk about design patterns, provenance, data sets for Linked Data on the Web

<OwenAmbur> Pat Franks will be compiling a similar report under the auspices of IBM's Center for the Business of Government

josema: kevin wanted me to make folks in the US aware of open government meetings about agency plans

<OwenAmbur> It would be nice if agencies posted their open gov plans in open StratML format

<OwenAmbur> If they fail to do so, I will probably convert some, if not all of them to StratML format myself

<Chris> @Owen - keep me posted on the IAC / Pats stuff - v interested

<josema> conference site -- http://www.tecnimap.es/

josema: the National Conference on EGovernment is next week, i will try to send information to the list

Discussion on data catalogues

<sandro> scovo http://sw.joanneum.at/scovo/schema.html

cygri: we at DERI are working with people in the UK for expressing statistics in rdf ; the Office of National Statistics in the UK did a workshop for how they should move forward with publishing their data in RDF, and explored using SCOVO and SDMX ; but have started working on an SDMX vocabulary for RDF
... there will be an early initial paper on this at ldow workshop

<josema> that will be very useful for us, too

John: this is work that I have commissioned ; once we get a good representation of sdmx for rdf is a key issue for Linked Data in the UK

cygri: i'm distributing my time between the dcat and sdmxrdf

josema: lets say you create that rdf vocab for sdmx, how will the relationship between the xml vocabulary and the rdf vocabulary work

cygri: sdmx started as an edi standard ; it includes an abstract model (uml), and there is an xml syntax
... since there is an abstract model, we are exploring how possible it would be to translate between the xml and rdf data ; there is a lot of statistical data that's not available as sdmx, so we're thinking about how to easily make that data available as sdmx rdf

John: jeni gave us an xslt demonstration of turning an sdmx xml document into rdf ... we figure the transformation of sdmx xml is going to be pretty straightforward
... there's also lgdx that is being used, and they have big systems that inhibit switching to something new ; so we want to be able to translate that as well
... once sdmxrdf becomes something we can use we can surface quite a bit of data quite quickly

<sandro> (nice timing)

John: we've started to use time interval uris too, which can be used to identify statistical data

josema: sounds like this deserves its own agenda item, some time in the future

<cygri> slides: http://richard.cyganiak.de/2010/03/dcat-for-egov-ig.pdf

josema: i would love to see an agenda item to talk only about this some time in the future

cygri: that's a link for some slides for the work so far

josema: ok the floor is yours

cygri: please follow along in the slides http://richard.cyganiak.de/2010/03/dcat-for-egov-ig.pdf
... most of the work has been by our egov unit and the linked data research center at deri

<josema> scribe: Ed Summers

cygri: it's about enabling the interoperability of government data catalogs we've seen popping up

<josema> scribeNick: edsu

going to talk about why it's important ; what's out there now ; what the dcat vocab is ; some experiments we've been doing and where we should go next

cygri: there are more than 30 government data catalogues online

<josema> compilation of the ones we found so far http://datos.fundacionctic.org/sandbox/catalog/ and http://datos.fundacionctic.org/sandbox/catalog/map ...

cygri: these efforts are done by both public and private parties

<josema> ... or in RDF as default SPARQL query at http://data.fundacionctic.org/sparql ;)

cygri: we're interested in this because the information is on the web and we'd like to query across the catalogs
... so information about san francisco can be found in local, state and federal catalogs
... in the EU we can see how me might want an eu level catalog ... data.gov.eu
... it would be nice to see new user interfaces for the data found in data catalogs ; which combine the metadata with the data itself ; also to do rating and social annotation of the data sets
... most of the data catalogs do export their data in a structured form
... however each has its own specific format, the documentation for it is lacking
... we did an in depth survey of 7 catalogs: some national some local
... we looked at the metadata, looking at the datasets themselves was out of scope
... we surveyed the types of metadata available
... quite a lot of metadata fields are shared, which is good news for interoperability
... we also looked at metadata fields, how consistently it was used, for example date fields
... we also looked at direct download links
... sometimes they go directly to the data, but quite often you go to a splash page, with a click through license, and find the download link on the page
... this is bad news for automatic processing of the data

<Chris> End

<josema> thx

cygri: the dcat vocab is at http://vocab.deri.ie/dcat ; and an overview is available at http://vocab.deri.ie/dcat-overview

<josema> all, queue yourself if you have questions and we'll go through them at the end, thx

cygri: we tried to keep in mind Hepp's Law: to be careful when designing a vocabulary not to make distinctions that aren't present in the data to be integrated

<josema> [that was very good advice]

cygri: we don't want to require data cleansing before dcat can be used
... rdf allows extensibility (classes and properties) to express additional information, so we focused on stuff that's in all the catalogs
... we tried to reuse from dublincore, skos and foaf, and minimized what we created ourselves
... we introduced dcat:Catalog, dcat:Dataset, dcat:CatalogRecord, dcat:Distribution
... Distribution is used indicate that format that the dataset is available in
... for example xml and json would be 2 Distributions
... categorization of datasets is common, and we used skos for this
... and the government agency that published the data was modeled with foaf:Organization
... we loaded some of the datasets into a standard relational database ; and then mapped to rdf with d2r http://lab.linkeddata.deri.ie/govcat/
... which has a sparql interface
... and we linked some things up to geonames and dbpedia
... for example we linked the agencies to dbpedia
... we did some example sparql queries for listing datasets that were published by agencies with budgets > a certain amount
... one of the benefits of using dcat is that it could enable distributed publishing, and federation later
... could also potentially allow datasets to be downloaded in an automated fashion
... also applications that worked on one data catalog could be repurposed
... we're looking for feedback on the vocabulary, and to get more eyes on the vocab
... we're writing up a guide to using dcat
... at deri we'd like to use dcat w/ voiD for describing rdf datasets
... also w/ sdmx+rdf ; metadata about the dataset is important
... what really has to happen is that it needs to happen not only at DERI but elsewhere on the web, by the catalog publishers

<John> This is very important work

cygri: one question is, as more people are involved how do we organize our work in a distributed way?
... to what extent would the egov interest group be a good place to do this?

<John> We need something like this for data.gov.uk RDF work

josema: thanks very much, we only have a few minutes for asking questions

Chris: have you done anything with the datacite consortium, and when you were looking at australia did you look at the australian data service?

cygri: no we just looked at data.gov au, would be interested in that

<sandro> PROPOSED: to extend the meeting, informally, for 15 minutes

John: i have someone looking at your work right now, would it be possible to version the document, right now it's still in draft

<josema> let's say 10, sandro ;)

cygri: we have to provide some documentation about the use of other vocabularies, we have good coverage for the new vocabulary we have introduced

John: do you need any help?

cygri: yes, always :-) perhaps we can discuss offline a bit
... it has to be driven by working on actual data

<josema> edsu: +

<John> +1

cygri: for v0.1 this is something we have to do here, it's hard to distribute

josema: we're going to extend the meeting by 10-15 minutes

<John> +1

sandro: this is very important work, i'm hoping that you can get users in the IG to help, but also institutional support from the w3c

<John> We should try and support somehow...

<josema> [I want to *thank* edsu for fantastic scribing today, wonderful, many many thanks]

<John> +1

cygri: i'm not sure if we should use the egov mailing list, or create another one (infrastructure) ; we need an issue tracker for this

<John> This is on the nail, practical, just what we need in UK

<josema> it's something we also need for our work at CTIC, happy to help from here, preferred if within W3C

cygri: an IG can publish notes at the w3c, for something like dcat to have more acceptance an IG Note would have a lot more acceptance

<John> And we can try and implement over next few weeks

sandro: another thing is the namespace for the vocab, perhaps a w3c namespace would lend it more credibility

John: it would for us

sandro: other applications i thought of: not every local gov't would have to make their own catalog, without requiring their own IT dept

<John> Yup!

<josema> +1

sandro: also, i was thinking it could be possible to publish mappings w/ dcat

<John> This is much much needed just now

josema: we really, really need this at my place of work
... sandro, how should we proceed?

<John> I'd love to see a group note on gov data catalogues

sandro: i'm happy to listen to what richard wants to do

cygri: i haven't followed too closely how the IG is being run with sub-projects
... with voiD we had good experience getting together a focused group, with a weekly call, with discussion list, issue tracker, and subversion repository
... i would try to replicate what we did w/ voiD

sandro: people seemed to want to use the egov ig discussion list for sub-projects
... maybe we could try to start on there with a [dcat] tag until someone complains

cygri: i think that could work

sandro: we could schedule a telecon, would richard be ok for chairing them?

cygri: yes, great

<josema> http://www.w3.org/egov/wiki/Projects

josema: that's the link fot the current projects

sandro: a lot of those aren't meeting so i wouldn't worry about it too much

<John> fabulous Richard!

+1

josema: next meeting aprl 14th
... we are adjourned

<josema> [ADJOURNED]

<josema> yup :)

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/03/31 14:45:32 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/we should/we should have/
Succeeded: s/moer/more/
Succeeded: s/to/too/
Succeeded: s/repositoy/repository/
Found Scribe: edsu
Inferring ScribeNick: edsu
Found Scribe: Ed Summers
Found ScribeNick: edsu
Scribes: edsu, Ed Summers
Default Present: George, Chris, +0203334aaaa, Sandro, +03498429aabb, edsu, cygri, josema, Owen?, John?, Vassilios
Present: George Chris +0203334aaaa Sandro +03498429aabb edsu cygri josema Owen? John? Vassilios
Regrets: kevin rachel cory
Found Date: 31 Mar 2010
Guessing minutes URL: http://www.w3.org/2010/03/31-egov-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]