Hello Lars, you can take a look at: http://toolserver.org/~kolossos/geo-export/ there you find the "missing-sv-coords.dat"-file (21MB) with the "missing" coordinates in Swedish wikipedia.
The other file in directory is the SQL-file to generate the list. I hope this help, it's the first shot and we can discuss about the format of the list.
Greetings Kolossos
Am 31.07.2011 00:01, schrieb Lars Aronsson:
On 07/30/2011 07:49 PM, Tim Alder wrote:
I found in my database 59.153 swedish articles[1] that could get coordinates from other wikipedias by interwikilinks. Compared with your number I see a potential of nearly 100%. In german Wikipedia we have 200.000 coordinates at 1.2 mio articles, this are nearly 20%. So there is a potential! (Also if such statistics are not 100% correct.)
Yes, I thought our current 8% is a bit low. I know that 29.6 % of all articles are biographies (5.6% women + 24.0 % men), because we categorize by gender just like the German Wikipedia, and I would guess that we could reach a similar number for articles tied to geographic coordinates. Interwiki would be a fine help for places outside of Sweden.
I have a bot account and I think I can inject the coordinates, but I don't have a toolserver account, so I need some help to compile the list. Maybe you can offer such lists for any language given as a parameter? The bot would have to adopt to local templates and policies, of course. Or the generated list could be used for manual projects. The MySQL statement could have a "limit 500", and after these have been fixed, the tool would generate the next 500.
Select count(*) from (Select distinct "T_sv" from "pub_C_geo_id" where "T_sv" is not NULL) as T