Data from Netherlands Statistics / CBS on Wikidata?

List overview All Threads
Download

newer

older

Wikidata API for fetching images...

New PHP library: Wikibase...

Arne Wossink

6 Aug 2015 6 Aug '15

8:41 a.m.

Dear list members, Wikimedia Netherlands has been approached by the CBS / Statistics Netherlands (https://en.wikipedia.org/wiki/Statistics_Netherlands), the governmental bureau for statistics, about a possible cooperation to make data available on a (yet to be determined) Wikimedia project. Since this is in a very preliminary stage, we don't have any further details at the moment, but I would be interested to hear if there could be any use of their data on Wikidata, or if similar data from similar institutions elsewhere have already been used. Best, Arne Wossink Projectleider / Project Lead Wikimedia Nederland Tel. +31 (0)6 11000505 *Postadres*: * Bezoekadres:* Postbus 167 Mariaplaats 3 3500 AD Utrecht Utrecht

Attachments:

attachment.htm (text/html — 1.6 KB)

Show replies by date

Gerard Meijssen

6 Aug 6 Aug

9:28 a.m.

Hoi, Data of a national statistics organisation is by definition TOO much. My recommendation would be to give it initial focus on something basic, something like the number of inhabitants of municipalities. Another thing that a statistics bureau is good for is for using it as a source. When data in Wikidata can be sourced to a statistics organisation, it would be wonderful to be able to link to the source and compare and update. When you start simple and focus on core functionality, it is easy to have it grow rapidly. I am happy to talk with the people of the CBS ... <grin> it helps that I speak Dutch and double Dutch </grin> thanks, GerardM On 6 August 2015 at 10:41, Arne Wossink <wossink(a)wikimedia.nl> wrote:

...

James Heald

1:39 p.m.

Too much to input as individual statements on a single item each with a historical date qualifier, I would agree. But there really ought to be a good way to store serial or tabular data on-wiki, with wiki-style version tracking, that could then eg be specified as source data for the new graph extension, or for inclusion into columns or tables, or accessible at the level of individual cells for incorporation into infoboxes. Are there steps that people have taken in the direction of wikiesque storage of such serial or tabular data ? (Sorry if this is a perennial question, but it's something I've been wondering for a while now -- all the better if there is a standard FAQ response that somebody could direct me to). All best, James. On 06/08/2015 10:28, Gerard Meijssen wrote:

...

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Daniel Kinzler

1:58 p.m.

Am 06.08.2015 um 15:39 schrieb James Heald:

...

Paul Houle

2:11 p.m.

So far as tables go, I would look at the data model behind XBRL. XBRL is sometimes maligned and often misunderstood, but essentially it is a format for presenting multi-dimensional tabular information. It is not really specific to finance, but financial reports really are hypercubes so it fits that domain. A wikified model would take more work, but it may be the way to go. One beef I have with time series data from the world bank is that many of them are piece-wise linear interpolated but don't document this. A geometric interpolation would be more plausible for one thing, but you see no bend in the curve caused by the '67 war in Nigeria, and no evidence that experts think the the eye-popping population numbers for Nigeria are a bit high. All of this matters little if you are doing a demo, but if you really care what the population of Nigeria is, this is a problem. On Thu, Aug 6, 2015 at 9:58 AM, Daniel Kinzler <daniel.kinzler(a)wikimedia.de> wrote:

...

Am 06.08.2015 um 15:39 schrieb James Heald:

Too much to input as individual statements on a single item each with a historical date qualifier, I would agree. But there really ought to be a good way to store serial or tabular data

on-wiki,

with wiki-style version tracking, that could then eg be specified as

source data

for the new graph extension, or for inclusion into columns or tables, or accessible at the level of individual cells for incorporation into

infoboxes.

Are there steps that people have taken in the direction of wikiesque

storage of

such serial or tabular data ? (Sorry if this is a perennial question, but it's something I've been

wondering

for a while now -- all the better if there is a standard FAQ response

that

somebody could direct me to).

We are thinking abotu it, but there are no concrete plans. We are really quite busy as it is already... -- Daniel Kinzler Senior Software Developer Wikimedia Deutschland Gesellschaft zur Förderung Freien Wissens e.V. _______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

-- Paul Houle *Applying Schemas for Natural Language Processing, Distributed Systems, Classification and Text Mining and Data Lakes* (607) 539 6254 paul.houle on Skype ontology2(a)gmail.com https://legalentityidentifier.info/lei/lookup/ <http://legalentityidentifier.info/lei/lookup/>

Thad Guidry

3:23 p.m.

Netherlands Statistics should just post the data on the web...so that anyone can use its "Linked Data". And actually, CSV on the Web is now a reality (no longer a need for XBRL) https://lists.w3.org/Archives/Public/public-vocabs/2015Jul/0016.html As DanBri notes in his P.S. at the bottom of the above link..."... the ability (in the csv2rdf doc) to map from rows in a table via templates into RDF triples....is very powerful." Thad +ThadGuidry <https://www.google.com/+ThadGuidry>

Gerard Meijssen

3:30 p.m.

Hoi, As far as I am concerned, data that is available on the web is fine if you use data on the web. It makes no difference when the data is to be used in the context of the WMF. When the CBS shares data with us in Wikidata, it makes the data available in Wikipedia. It is why I would like for something small, a pilot project something where we can build on. Thanks, GerardM On 6 August 2015 at 17:23, Thad Guidry <thadguidry(a)gmail.com> wrote:

...

Jane Darnell

4 p.m.

I have used the CBS website to compile my own statistics for research. Their data is completely available online as far as I know and you can download the queries you run on the fly in .csv file format, or text or excel. They have various data tables depending on what you find interesting and complete tables of historical data is also available. That said, I think any pilot project would need to start with their publications, which are also available online. These can be freely used as sources for statements. Interesting data for Wikidata could be population statistics of major cities per century or employment statistics per city per century and so forth. See CBS.nl On Thu, Aug 6, 2015 at 5:30 PM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Joe Filceolaire

7 Aug 7 Aug

1:06 p.m.

...

wrote:

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

John Erling Blad

8 Aug 8 Aug

11:10 a.m.

There have been some discussions about reuse of statistics with people from Statistics Norway.[1] They use a format called JSON-stat.[2] A bunch of census bureaus are starting to use JSON-stat, for example Statistics Norway, UK’s Office for National Statistics, Statistics Sweden, Statistics Denmark, Instituto Galego de Estatística, and Central Statistics Office of Ireland. I've heard about other too. I have started on some rant at Meta about it, I didn't finish it.[3] Perhaps more people will join in? ;) A central problem is that statistics are often produced as a multidimensional dataset, where our topics are only single indices on one of the dimensions. We can extract the relevant data, but it is probably better to make a kind of composite key into the dataset to identify the relevant stuff about our topic. That key can be stored as a table-specific statement in Wikidata, and with a little bit of planning it can be statistics-specific or even bureau-specific. [1] https://ssb.no/en/ [2] http://json-stat.org/ [3] https://meta.wikimedia.org/wiki/Grants:IdeaLab/Import_and_visualize_census_… On Fri, Aug 7, 2015 at 3:06 PM, Joe Filceolaire <filceolaire(a)gmail.com> wrote:

...

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Arne Wossink

13 Aug 13 Aug

7:56 a.m.

Thanks all for the input! I would like to stress again that the first meeting with the CBS has yet to take place, but we will definitely take the views expressed here into account. Best, Arne Wossink Projectleider / Project Lead Wikimedia Nederland Tel. +31 (0)6 11000505 *Postadres*: * Bezoekadres:* Postbus 167 Mariaplaats 3 3500 AD Utrecht Utrecht 2015-08-08 13:10 GMT+02:00 John Erling Blad <jeblad(a)gmail.com>om>:

...

Hypercubes and csv flat files belong in commons in my opinion (commons

may

have a different opinion ). That's if we even want to store a copy. This source data should then be translated into wikidata triples and statements and imported into wikidata items. The statements in wikidata are then used to generate lists and tables and graphs and info graphics in wikipedia. At least that's how I see it Joe On Thu, 6 Aug 2015 17:00 Jane Darnell <jane023(a)gmail.com> wrote: > > I have used the CBS website to compile my own statistics for research. > Their data is completely available online as far as I know and you can > download the queries you run on the fly in .csv file format, or text or > excel. They have various data tables depending on what you find

interesting

> and complete tables of historical data is also available. That said, I

think

> any pilot project would need to start with their publications, which are > also available online. These can be freely used as sources for

statements.

> Interesting data for Wikidata could be population statistics of major

cities

> per century or employment statistics per city per century and so forth.

See

> CBS.nl > > On Thu, Aug 6, 2015 at 5:30 PM, Gerard Meijssen > <gerard.meijssen(a)gmail.com> wrote: >> >> Hoi, >> As far as I am concerned, data that is available on the web is fine if >> you use data on the web. It makes no difference when the data is to be

used

>> in the context of the WMF. >> >> When the CBS shares data with us in Wikidata, it makes the data

available

in Wikipedia. It is why I would like for something small, a pilot project something where we can build on. Thanks, GerardM On 6 August 2015 at 17:23, Thad Guidry <thadguidry(a)gmail.com> wrote: > > Netherlands Statistics should just post the data on the web...so that > anyone can use its "Linked Data". > > And actually, CSV on the Web is now a reality (no longer a need for > XBRL) > > https://lists.w3.org/Archives/Public/public-vocabs/2015Jul/0016.html > > As DanBri notes in his P.S. at the bottom of the above link..."... the > ability (in the csv2rdf doc) to map from > > rows in a table via templates into RDF triples....is very powerful." > > Thad > +ThadGuidry > > _______________________________________________ > Wikidata mailing list > Wikidata(a)lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/wikidata > _______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Gerard Meijssen

9:10 a.m.

Hoi, Make sure that you take a Wikidata person with you. It makes a difference. Maarten, Sjoerd, Jane, Edo, me.... they know the project and what we can do. They will help them get their feet wet and not drown in all the technical nitbits. Thanks, GerardM On 13 August 2015 at 09:56, Arne Wossink <wossink(a)wikimedia.nl> wrote:

...

Hypercubes and csv flat files belong in commons in my opinion (commons

may

and

graphs and info graphics in wikipedia. At least that's how I see it Joe On Thu, 6 Aug 2015 17:00 Jane Darnell <jane023(a)gmail.com> wrote: > > I have used the CBS website to compile my own statistics for research. > Their data is completely available online as far as I know and you can > download the queries you run on the fly in .csv file format, or text or > excel. They have various data tables depending on what you find

interesting

> and complete tables of historical data is also available. That said, I

think

> any pilot project would need to start with their publications, which

are

> also available online. These can be freely used as sources for

statements.

> Interesting data for Wikidata could be population statistics of major

cities

> per century or employment statistics per city per century and so

forth. See

be used

>> in the context of the WMF. >> >> When the CBS shares data with us in Wikidata, it makes the data

available

>> in Wikipedia. >> >> It is why I would like for something small, a pilot project something >> where we can build on. >> Thanks, >> GerardM >> >> On 6 August 2015 at 17:23, Thad Guidry <thadguidry(a)gmail.com> wrote: >>> >>> Netherlands Statistics should just post the data on the web...so that >>> anyone can use its "Linked Data". >>> >>> And actually, CSV on the Web is now a reality (no longer a need for >>> XBRL) >>> >>> https://lists.w3.org/Archives/Public/public-vocabs/2015Jul/0016.html >>> >>> As DanBri notes in his P.S. at the bottom of the above link..."...

the

>> ability (in the csv2rdf doc) to map from >> >> rows in a table via templates into RDF triples....is very powerful." >> >> Thad >> +ThadGuidry >> >> _______________________________________________ >> Wikidata mailing list >> Wikidata(a)lists.wikimedia.org >> https://lists.wikimedia.org/mailman/listinfo/wikidata >> > > > _______________________________________________ > Wikidata mailing list > Wikidata(a)lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/wikidata > _______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Sjoerd de Bruin

11:13 a.m.

+1, make sure to bring someone from the project with you. Sjoerd de Bruin

...

Op 13 aug. 2015 om 11:10 heeft Gerard Meijssen <gerard.meijssen(a)gmail.com> het volgende geschreven: Hoi, Make sure that you take a Wikidata person with you. It makes a difference. Maarten, Sjoerd, Jane, Edo, me.... they know the project and what we can do. They will help them get their feet wet and not drown in all the technical nitbits. Thanks, GerardM

On 13 August 2015 at 09:56, Arne Wossink <wossink(a)wikimedia.nl> wrote: Thanks all for the input! I would like to stress again that the first meeting with the CBS has yet to take place, but we will definitely take the views expressed here into account. Best, Arne Wossink Projectleider / Project Lead Wikimedia Nederland Tel. +31 (0)6 11000505 Postadres: Bezoekadres: Postbus 167 Mariaplaats 3 3500 AD Utrecht Utrecht 2015-08-08 13:10 GMT+02:00 John Erling Blad <jeblad(a)gmail.com>om>:

Hypercubes and csv flat files belong in commons in my opinion (commons may have a different opinion ). That's if we even want to store a copy. This source data should then be translated into wikidata triples and statements and imported into wikidata items. The statements in wikidata are then used to generate lists and tables and graphs and info graphics in wikipedia. At least that's how I see it Joe On Thu, 6 Aug 2015 17:00 Jane Darnell <jane023(a)gmail.com> wrote: > > I have used the CBS website to compile my own statistics for research. > Their data is completely available online as far as I know and you can > download the queries you run on the fly in .csv file format, or text or > excel. They have various data tables depending on what you find interesting > and complete tables of historical data is also available. That said, I think > any pilot project would need to start with their publications, which are > also available online. These can be freely used as sources for statements. > Interesting data for Wikidata could be population statistics of major cities > per century or employment statistics per city per century and so forth. See > CBS.nl > > On Thu, Aug 6, 2015 at 5:30 PM, Gerard Meijssen > <gerard.meijssen(a)gmail.com> wrote: >> >> Hoi, >> As far as I am concerned, data that is available on the web is fine if >> you use data on the web. It makes no difference when the data is to be used >> in the context of the WMF. >> >> When the CBS shares data with us in Wikidata, it makes the data available >> in Wikipedia. >> >> It is why I would like for something small, a pilot project something >> where we can build on. >> Thanks, >> GerardM >> >> On 6 August 2015 at 17:23, Thad Guidry <thadguidry(a)gmail.com> wrote: >>> >>> Netherlands Statistics should just post the data on the web...so that >>> anyone can use its "Linked Data". >>> >>> And actually, CSV on the Web is now a reality (no longer a need for >>> XBRL) >>> >>> https://lists.w3.org/Archives/Public/public-vocabs/2015Jul/0016.html >>> >>> As DanBri notes in his P.S. at the bottom of the above link..."... the >>> ability (in the csv2rdf doc) to map from >>> >>> rows in a table via templates into RDF triples....is very powerful." >>> >>> Thad >>> +ThadGuidry >>> >>> _______________________________________________ >>> Wikidata mailing list >>> Wikidata(a)lists.wikimedia.org >>> https://lists.wikimedia.org/mailman/listinfo/wikidata >>> >> >> >> _______________________________________________ >> Wikidata mailing list >> Wikidata(a)lists.wikimedia.org >> https://lists.wikimedia.org/mailman/listinfo/wikidata >> > > _______________________________________________ > Wikidata mailing list > Wikidata(a)lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/wikidata _______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

_______________________________________________ Wikidata mailing list Wikidata(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata

Peter F. Patel-Schneider

6 Aug 6 Aug

3:59 p.m.

I am rather puzzled by some of the attitudes displayed in this discussion. It would have thought that information from Netherlands Statistics meets the requirements for inclusion in Wikidata. It is a primary source of high-quality, notable information. Further much of the information is related to existing Wiki... pages. Each time I consider this I think "Wouldn't it be nice if [some bit of information from Netherlands Statistics] was available on [some Wiki... page]?" Wouldn't it be nice if the populations of all Netherlands populated places in Wikipedia automatically had the most-current population information sourced from Netherlands statistics? Wouldn't it be nice if summary economic information about the Netherlands was similarly automatically updated? Of course, a national statistical office gathers a lot of data, and much of it may not be of general interest. So, perhaps, not all of the Netherlands Statistics information should be automatically imported into Wikidata. Maybe instead there should be a way for this information to be accessible via Wikidata so that it can be easily used when creating Wiki... pages. How to make this work? Good question. There are a number of issues that have to be worked out, including resource consumption issues, user interface issues, and information organization issues that go beyond linked data issues. Maybe Netherlands Statistics can help, particularly if the Wikidata community was enthusiastic about collaborating with them instead of just again saying "Too much information". Peter F. Patel-Schneider Nuance Communications

3179

days inactive

3186

days old

wikidata@lists.wikimedia.org

Manage subscription

13 comments

11 participants

tags (0)

participants (11)

Arne Wossink
Daniel Kinzler
Gerard Meijssen
James Heald
Jane Darnell
Joe Filceolaire
John Erling Blad
Paul Houle
Peter F. Patel-Schneider
Sjoerd de Bruin
Thad Guidry