W3C

eGov IG Meeting (Eurasian)

21 Jan 2013

Agenda

See also: IRC log

Attendees

Present
+88697069aaaa, +8532871aabb, +8532871aacc, +1.703.281.aadd, PhilA2, thschee, +3539149aaee, Brand, florian, Deirdre, Keitha, [IPcaller], HadleyBeeman
Chair
Tomasz Janowski
Scribe
Florian Henning, fhenning, PhilA2

Contents


<PhilA2> tomasz: Opens the meeting

tomasz: welcomes everyone, introduces speakers
... asks for tour de table for personal introductions

phil, i will be scribing

<PhilA2> scribe: fhenning

<PhilA2> scribeNick: fhenning

<SamG> fhenning is scribing ... i am learning how to scribe

thanks phil

Open Government Information and Data in New Zealand (Keitha Booth)

<PhilA2> slides are at http://www.w3.org/egov/wiki/images/2/24/WC3_egovt%2C_21_January_2013.pdf

Keitha_: slide2:
... ...Clear Cabinet decision and intent “proactive disclosure” OIA review

phil, i'm sorry, audio is again too bad here - may i ask you to take over with scribing?

thank you

<PhilA2> slide 5 shows layers and importance of various factors

<PhilA2> slide 6

<PhilA2> Keitha: NZ Gov retains copyright

<PhilA2> ... uses Creative COmmons, mostly ccBy

<PhilA2> ... data must be published on data.govt.nz. Praised by Andrew Stott but he also noted that it's not being used enough

<PhilA2> ... features include requests for data

<PhilA2> Reference point for those unfamiliar with Andrew Stott, see bottom of page at http://www.cabinetoffice.gov.uk/content/public-sector-transparency-board-who's-who

<PhilA2> slide 9 (I think)

<PhilA2> Keitha_: emphasises use of CSV as open format

<PhilA2> Keitha: Highlights some users of NZ data, such as Zoodle

<PhilA2> ... watch my street is a competitor

<PhilA2> ... koordinates hosts a lot of topographical info. used a lot by government itself (scribe I think I caught that right)

<PhilA2> scribe: PhilA2

<scribe> scribeNick: PhilA2

Keitha: Good examples on unintended uses of the data

slide 11 shows more examples of use of the data

slide 14 shows uses of the data for transparency

slide 15 - should gov be producing Apps or should third parties do that?

schools have found creating Apps a useful way to get info to people

scribe: also holiday makers. Lots of Apps for tide tables, tourist-related Apps etc. Often collaborations between gov departments and local groups

slide 16 - talking about sustaining the work

Keitha_: Working on strategy for 2017. AN update of current eGov strategies. Managing data as a strategic asset is an important part of that

PhilA2: What is reasonably priced?

<GwynSutherlin> I thought that was a great slide

Keitha: We'd like it to be free, but we're not able to go as far as free. It's the cost of dissemination

PhilA2: Can you measure the economic benefit oft he data? and can that show that the revenue more than covers the cost of publication?

Keitha: Looked at Uk and Aus. The normal cost-benefit analysis won't work. Advised to focus on higher level economic impacts

<laurenceNZ> I think it is more the economic benefit rather than revenue

Keitha: we're now looking at prepaing case studies

<SamG> sorry just exploring

Keitha_: Looking ahead to analysis in 3 years' time to whether we have met the target for the system as a whole

tomasz: Is there some systematic way to measure impact on gov and priv sector?

Keitha: Not at the moment. We're relying on what departmetns are telling us
... Had an early benchmark report but it was too early to give us an answer on that. But I'm happy to report back in 6 months' time
... Expecting to get a report to cabinet in that period

<SamG> how is this OIA?

tomasz: Thanks Keitha for the presentation

From 0 to 5 Cities - Open Data in Taiwan (TH Schee)

<GwynSutherlin> thanks Keitha, very interesting

Slides are at http://www.w3.org/egov/wiki/images/2/26/W3c-egov-opendatatw.pdf

<laurenceNZ> There was an EU case study that showed a 1.7%impact on GDP from open data

<laurenceNZ> Links to work from the EU that gives some teeth to the economic value argument. The headline is "direct and indirect economic benefits for the economy are of the order of 1.7% of GDP" .. which should be a compelling story. LINKS: the research http://bit.ly/MJFd9X the powerpoint http://bit.ly/N20GeZ Also some work, led by Deloitte, into the pricing of PSI: the powerpoint http://bit.ly/M7VvNO the summary report (2MB) http://bit.ly/KIti1N full repo[CUT]

thschee: We have a lot of data published on the RoC site
... Translating a lot of info from OKFN and elsewhere. Open Data movement growing in Taiwan

<scribe> ... new open data legislation being drafted

UNKNOWN_SPEAKER: regulation due to take effect in a matter of months
... Open Data portal opened last year
... you can find a lot of cultural data there

PhilA2: Notes that Wang Mei-Hsueh, Academia Sinica Research Center for Information Technology Innovation, Taiwan took part in last year's W3C workshop on this (Deirdre and GwynSutherlin were there ;-) http://www.w3.org/2012/06/pmod/agenda

thschee: Portals like data.gov.tw often launched in a bit of a rush - c'est la vie
... We have 3 portals but licensing not consistent, not all free and so on so developers are having a hard time getting the data.
... using Microsoft's OGDI platform

PhilA2: Reference for Microsoft's OGDI http://www.microsoft.com/government/ww/public-services/initiatives/Pages/open-government-data-initiative.aspx

thschee: Monitoring usage of Apps so we can see the demand - which is high
... gov promoting Apps
... if you are creating Apps using the Taiwan open data - you;re doing better than the gov wehich is finding it a little hard

slide 6

thschee: Getting a lot of stakeholders together
... Creative Commons Taiwan
... Open Street Map Taiwan etc.
... some people doing better job of publishing gov data by scraping Web sites than gov itself doing by publishing the data
... common belief that your data will be used by others to do you harm

slide 7 - focus on current strategy

slide 8 - shows relationships. Taiwan has hackathons every month

scribe: ecosystem is cooperative effort in that it's gov and private sector

PhilA2: Can you expand on the issue of people being reluctant to publish their data through fear

thschee: We're working to expose the work and get it on the agenda of politicians through the media
... then you have to deliver solutions that may be open source-based
... with launch of open data portals, lots of companies behind gov tenders, some thought it would be a good cash cow...
... needs high level support - work top down

PhilA2: So with talk about open source and companies circling round looking for contacts - why go for Microsoft rather than, for e.g. CKAN?

<BRAND> I have found that some commercial products are more mature for data federation than CKAN

PhilA2: confidence, speed, and CKAN not good witch Chinese characters

thschee: gov wanted a quick solution that works out of the box

tomasz: On your last slide - do you have a concept of maturity in the use of open gov data?
... how mature is the use of OGD?

thschee: The tech start ups are pretty mature. They know about the limitations, esp. of licnces etc.
... making good profits with small teams of developers

Research on US Federal Government Handling of Data (Brand Niemann)

slides http://www.w3.org/egov/wiki/images/4/47/BrandNiemann01212013.pdf

<BRAND> http://semanticommunity.info/@api/deki/files/21031/OpenGovernmentDatainJapan01182013.pdf Additional backup slides for Brand Niemann

BRAND: introduces himself and his work
... been asked to help the Data Transparency Coalition (see recent talk by Hudson Hollister)
... working with big dataand analytics
... helping open data for Japan
... getting the right people, etc. in place
... also working on strategy for big data more than open data
... now in era of data scientist

Slide 2

Been asked to look at current landscape of open data round the world.

Look at Mosaic effect where multiple sources of data together reveal personal or otehr data that might be detrimental to national security

slide 3 shows current big picture

slide 4 offers a big solution

BRAND: People I speak to are all supportive but they're looking for ROI. We need something soon or we'll move on to something else
... talks about cell level secutiry

<laurenceNZ> Laurence is leaving the call

BRAND: We already have a good deal of knowledge about open government data. Refers to recent NIST conference on cloud copmputing
... someone from intelligence community looking at potential mosaic effect, someone with good background in data analytics

slide 7. NIEM admits not had as much success with working with private sector as they'd like. Oracle prob had more success at implementing NIEM than NIEM itself

slide 8 - METI - biggest ministry in Japan

BRAND: Govs need chief data officers, they need teams of data scientists and statisticians...
... they need a return on investment. I've been saying they should work with the right people/skills
... I've developed a 5 step recipe (slide 10)
... Spotfire dashboard seems more stable and powerful than, for example, CKAN

5 steps are independent of the tools (I've just shown the ones I use)

slide 11 maps this to TimBLs 5 stars data

slide 12 shows the Japanese yearbook in the dashboard. Takes a while to load but very powerful

scribe: individual data sets are small but combined together they form big data
... I was interested in the recent earthquake and its effects

slide 14 - some of the data in CKAN portal is not to well sourced as others

Todd Park's deputy reckons that city data is prob more valuable than the national stuff

BRAND: Conclusion (slide 16)

Keitha: What is the definition of statistics. In NZ we're moving in the open data world - the stuff they have is mostly survey results cf. administrative data - stuff that doesn't come from surveys

BRAND: We had a recent conference with the Association of Public Data users. The people from that stat community (framework data) all realisedd that there's this new source of data from social media
... it's scary as it's not collected methodically, But they recoignise it's improtant
... meanwhile the social media people are becoming aware of the otehr community
... we need to brign the two together of course. It's exciting to work with both
... the proceedings of that are interesting. People from statistical agencies looking at how social media data can be brought in in such a way that the stat agencies can use

Todd Park has said that data scientists need to come to our hackathons and expalin what this data means

Keitha: Let's keep talking. The otehr area for us is anonymisation of personal data

<SamG> how can you validate social media data?

Keitha: you talked about security etc. I'd like to continue talking about that with you. Tends to be about anonymisation of survey datam rather than admin data

BRAND: People who work on the data are sworn to lie - i.e. not reveal what they've seen - and then they aggregate up until it becomes anonymous

<BRAND> http://semanticommunity.info/AOL_Government/APDU

tomasz: Closes the meeting

<HadleyBeeman> Thanks, tomasz!

<GwynSutherlin> thanks

tomasz: Next meeting topic is open to suggestion. We've had 5 on this topic