Many organizations that collect statistics on a regular basis document their actual datatype, their methods for collecting the statistics, how they process the statistics, how they present them and how they validate and invalidates them. Its often just called the "metadata", but the different parts are quite distinct and not really metadata. Its also worth noting that specific entries can change role, and in some cases data used in some role becomes metadata. At Statistics Norway the information about how the data was collected was called the "vaskeseddel" ("laundry note" http://en.wikipedia.org/wiki/Laundry_symbol). I think we should find some simplified method to collect and make available this information. John
On Fri, Mar 15, 2013 at 8:49 AM, Mathieu Stumpf psychoslave@culture-libre.org wrote:
Le 2013-03-14 19:38, Michael Hale a écrit :
A topic I've been involved in recently regards statistics for gun violence in the US. The government publishes a big report every year, but it takes them most of the year to collect the information from all of the local police agencies and compile the results. Several English Wikipedia articles use this information, and it would be awesome if the tables in the articles could be generated automatically from data in Wikidata. It seems like ideally I would have some code I would run whenever they release the new report that would automatically import all of the data into Wikidata and add the appropriate references. I suppose the information would go in the item for each city. Say for the Atlanta item, there would be a statement for murders and the value would be a number and the qualifier for these statements would just be "2011" or whatever. Then I would want to be able to have a template that automatically makes a table to show the 5 most recent years somewhere in the Atlanta article for example.
That's great. Now we should also provide with each statiscal generated information an explanation of how it was interpreted. Numbers aren't as objective as one may believe, so we should take care to explain methodologies we use.
kind regards, mathieu
Date: Thu, 14 Mar 2013 16:37:01 +0100 From: jeblad@gmail.com To: wikidata-l@lists.wikimedia.org Subject: Re: [Wikidata-l] Expiration date for data
This has also been aired in other discussions. Outdated entries can both be something that is only valid within a set timeframe, but can also be dependent on something else. One special case is when an external source do not support a specific statement anymore.
On Sun, Sep 30, 2012 at 11:59 AM, Marco Fleckinger marco.fleckinger@gmail.com wrote:
Hi,
regarding an actual topic in Germany about publication of the timetable-data of Deutsche Bahn (German national railway company) and their willingness of a discussion with other Open-Data-Supporters it may be a good idea of providing an expiration dates for Wikidata-records.
In their open letter to Mr. Kreil [1] they announced that it may cause problems providing the timetable-data in an open way if e.g. anybody uses old data.
Marco
[1] http://www.db-vertrieb.com/db_vertrieb/view/service/open_plan_b.shtml
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
-- Association Culture-Libre http://www.culture-libre.org/
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l