Linked Data/Gov Publishing

31 Oct 2012

See also: IRC log


Bernadette Hyland
eric, ericP


<sandro> scribe: eric

<FabGandon> scribe:ericP

Armin: w3c office OZ
... interested in SemWeb, part of linked data platform WG
... started OZ egov data WG last week

Gabriel Kepeklian: Societe ATOS in FR. datalift project

jaques: fujitsu. closest to linked data is as a heavy XML (XSLT) practitioner

RalphS: hear for use cases

<SteveS> Hi, I'm Steve Speicher with IBM, working on Linked Data Platform WG…interested in use cases and how LDP can apply

ching...: LD taiwan

bert bos: W3C staff

eric V: evaluation task force. interested in anything with accessibility

felix sasaki: DFKI in DE

<FabGandon> Fabien Gandon, in charge of Wimmics Lab at Inria, http://wimmics.inria.fr interested in reconciling social semantics and formal semantics in web applications, linked data, semantic web, ontologies, working in the datalift project, http://datalift.org/

Bernadette: would like to see the W3C be more active in LD
... eGov is very broad. two co-chairs: Tomaz, Jeanne Holm

<JeniT> Jeni Tennison, Technical Director at Open Data Institute, interested in publication and consumption of open data, which may be linked data

[Bernadette:'s slides]

Bernadette: 15km view: aiming to create a sustainable workflow around capture and publication of government LD
... many gov execs talk about the importance of open data.
... newling talking about machine-readable

[slide: growing chorus]

Bernadette: enabled with standards and agencies publishing 5* linked data
... moving from research orgs, e.g. DERI, Gent University
... i feel that W3C needs to talk about how LD fits into the big data landscape
... we have editors in our WG who write vocabs, which already have many implentations
... WG has civil employees and practionres
... alternate between vocabs and best practices
... three vocabs in REC-track

<SteveS> Note the Linked "Open" Data mug has been updated to include the work "Open" http://www.cafepress.com/w3c_shop.597992118

Bernadette: making docs available on web platform

<SteveS> for 5 star Linked Open Data

Bernadette: some vendors are marketing linked data for other purposes
... w3c needs to clarify the brand
... would like to hear W3C strengths and weakness wrt govornment outreach

@gent: need to focus on machine-readable

scribe: everyone can participate -- need to emphasize this

SteveS: work on in integration for not just production but consumptiopn and manipulation

BartvanLeeuwen: govornments oftent reluctant to put data on the web but others can make use of it

<ahaller2> +1

HadleyBeeman: we see e.g. two local councils trying to publish data but not working together
... focus on transfering data between systems rather than APIs

FabGandon: govs have data that they publish, but also need to use the data internally

Bernadette: how could W3C do more to further your efforts? what do you need?
... .. marcom?
... .. specs?
... .. conversations with senior officials?

Armin: missing best practices for govts
... want to know how subdomains should use vocabs

Bernadette: (as lead author):
... .. how to mint IRIs
... .. cookbook
... there's a book called "Linking Government Data"

ivan: when you say "W3C", do you mean the 60 folks who work at w3c or the working groups and the community?
... i see the W3C folks as being facilitators for the community
... when *i* talk about LD to outsiders (as devil's advocate), i'm missing the convincing implementations which must be done with LD
... some exists, e.g. BartvanLeeuwen's, but we don't have the full persuasive power

<FabGandon> ericP: I work in healthcare a lot

<FabGandon> ... if I come with a cool RDF application it can allways be done another way

<FabGandon> ... if we want to convince people we need a business analysis

<sandro> eric: Anything you build with Linked Data you can build without it, but doing it with RDF has certain advantages -- easier, faster, extensible, connecting across domains. But we need a business analysis, where people do it BOTH ways, and report both costs.

<FabGandon> ... a comparison between two ways of doing things

<FabGandon> ... this the way we will influence the way the budget will be invested

<FabGandon> ... otherwise it is hard to sale the case

<RalphS> ericP: ... Another way. Our intuition is that Linked Data is the most extensible way

HadleyBeeman: to ericP and ivan, pointing to business case is fantastic, and we mostly see potential more than success stories
... i sat down with JeniT to see what happens when we *don't* use RDF
... most of our datasets are small and incomplete

<FabGandon> btw my point was that Linked Enterprise Data also applies to Governement Data, i.e. there is a case for none open data within the firewalls of government bodies.

HadleyBeeman: e.g. two columns of pothole location and a road name
... being able to annotate gives that actual value

<sandro> HadleyBeeman: lots of data like two column list of potholes -- date + road name. Connect with post code or something and you know the location. We still have the problem that a lot of our developers dont want to deal wit RDF.

HadleyBeeman: of course our developers don't wnat to deal with RDF, so we write lots of transforms

shoaib: we have lots of govt clients. we need to express the value

Bernadette: everything should be driven by use cases. i'd be happy if we had a place to write up case studies
... we no longer need to search for use cases in the wild.
... just came up with three from 3 round stones

ivan: some years ago we had Semantic Web Education and Outreach group
... intially, lots of people published their case studies, btu it trickled down
... without a consistent manager, this info deteriorates

Bernadette: one of the WG deliverables was a community directory
... people put info out as if it's a product, but as an academic project, it can go away at any time
... the interface on http://dir.w3.org/ needs work, but it still can serve as a community directory
... the ROI figures often don't exist, and won't for a few years
... knowing that there are other cities investing in LD helps them write the check

<sandro> (Surely we're still among Innovators, not Early Majority. That was for html5.)

Armin: this WG should be a facilitator for the governments serving the public good

<sandro> HadleyBeeman: tracking ROI costs additional money

Armin: the intial invenstors will always get a negative ROI

HadleyBeeman: tracking ROI is expensive; hard when budgets are tight

BartvanLeeuwen: success stories often focus on agility. folks don't know where they're going and leverage simple steps

JeniT: i don't talk about RDF at all
... i don't even talk about LD. just URIs for things and opening things up

<sandro> HadleyBeeman: govt (technology poliicy) is 80% politics and emotions

<sandro> JeniT: People put stuff up for the public good, but that wont keep it up. For that, they need ongoing benefit.

JeniT: re: ROI for govt. govt can say they're acting in public good, but that's not sufficient
... need to focus on ongoing benefits

judy: do you have a set of plain language talking points?
... could be useful in the egov part of the W3C website
... in WAI, folks asked us for a business case.
... rationale for invensting in accessibility wouldn't work for all business and orgs
... we now have set of four customizable use cases

<JeniT> technical factors / legal/regulatory / social responsibility / economic factors

judy: finanical benefit, social responsibility, legal and regulatory, technichal benefits (later re-use)

Bernadette: working with agencies, we need to help the procurment officer
... i'm trying to sell health care companies on the use of gov't data
... they look bad if the e.g. FDA feed goes out of date

JeniT: lots of data sites say "there is no guarantee this will persist"
... business can't build on that
... you need a formal contract with the suppliers

Bernadette: like an MOU

JeniT: we've signed MOUs with serveral businesses

ivan: is there a technical solution?

<sandro> I believe we're 5 minutes over time....

ivan: our model is that a govt agency publishes the data and presumes someone will use it
... if they knew the consumer, maybe they'd be more careful

martinAlvarez: training is important. developers need to see the value of using LD

<Judy> An example of a customizable business case for take-up of standards,

<Judy> from W3C Web Accessibility Initiative (WAI) Education and Outreach WG

<Judy> "Developing a Web Accessibility Business Case for Your Organization"

<Judy> and then see the four "factors" areas highlighted under the top page, at: http://www.w3.org/WAI/bcase/

martinAlvarez: the UK govt provides such training
... in my experience, it's easy to convince govts to use LD to relase public data, but they don't see others using the data

<Ruben> rssagent, please make logs world-visible

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.137 (CVS log)
$Date: 2012/10/31 15:06:44 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.137  of Date: 2012/09/20 20:19:01  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Jean Holme/Jeanne Holm/
Succeeded: s/gabriel: ethos/Gabriel Kepeklian: Societe ATOS/
Found Scribe: eric
Found Scribe: ericP
Inferring ScribeNick: ericP
Scribes: eric, ericP

WARNING: No "Topic:" lines found.

Present: FabGandon

WARNING: Fewer than 3 people found for Present list!

Got date from IRC log name: 31 Oct 2012
Guessing minutes URL: http://www.w3.org/2012/10/31-linkeddata-minutes.html
People with action items: 

WARNING: No "Topic: ..." lines found!  
Resulting HTML may have an empty (invalid) <ol>...</ol>.

Explanation: "Topic: ..." lines are used to indicate the start of 
new discussion topics or agenda items, such as:
<dbooth> Topic: Review of Amy's report

[End of scribe.perl diagnostic output]