[OSM-talk] Wikidata accuracy (Was: MAPS.ME edits - partly sub-standard)

[Martin Koppenhoefer]

sent from a phone

> Il giorno 26 ago 2016, alle ore 13:51, Dave F <davefoxfac63 at btinternet.com> ha scritto:
> 
> I'm surprised Wikidata is different to Wikipedia. I thought the former was derived from the latter.
> (btw The 7,172,036 figure isn't the latest data in Wikidata) 


oops, sorry, you are right, there's a 2012 estimate in wikidata (8,416,535), but the 2011 census is missing, Wikipedia/en (London) has a mid 2015 estimate.

in osm we could decide to use the population of the urban area (built up, former "urban area") on the place object (especially if it's a node): https://en.wikipedia.org/wiki/Greater_London_Built-up_Area
and http://www.nomisweb.co.uk/articles/747.aspx
the other figure could go on the administrative boundary object.


what I find quite strange in wikidata: both, the capital object from the main article
https://m.wikidata.org/wiki/Q84
and the London Commuter Belt
https://m.wikidata.org/wiki/Q2422792
are marked as instances of "Metropolitan Area"s but the population figures for the former are referring to the city, not to the metropolitan area, and the latter has no population info, unlike Wikipedia which quotes Eurostat with provisional 14,031,830 for 2014. I propose to remove the metropolitan area instance relation from q84.

There's also Greater London which is here in wikidata https://m.wikidata.org/wiki/Q23306 and had a population of 8,308,369  in 2012, likely in Osm this can be found or puzzled together from administrative entities (did not check).

cheers,
Martin