[OSM-talk] How can we measure the completeness of the map statistically

[Mark Williams]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

wer-ist-roger wrote:
> Am Donnerstag 26 Juni 2008 schrieb Frederik Ramm:
>>>     * Another hypothesis is that more complete areas of OSM will  
>>> have a higher level of edit activity.
>> Then there are those who say that an area that's complete doesn't  
>> require any edit activity...
> 
> That is right that "complete" areas might not gany much edits (or maybe no 
> edits at all) but you can see that on the edit history you will get a 
> submit-curve that has a lot of edits within a timeperiod until you get at 
> thet point it's getting less end less till you have almost no edits at all.
> 
> So a calculation should allways include the history of submitted changes.
> 

Indeed, that curve of much editing -> less editing should be a factor in
the completeness measure - an area with much activity in the past + some
more recent would be good, an area with no edits at all for 2 years is
ripe for a check-over, an area with no history is clearly not good.
Landuse=field is perhaps an example of the exception that proves the rule?

I still think it would be useful for an overlay to exist for mappers to
mark up where they consider done, perhaps that & an auto-layer should
combine to make a more visible 3rd layer of completeness where they
agree? We could start by filling this public layer with coastline data &
mark the seas as done, at least - unless we want to get full marine data
in as well?

Mark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFIZmF7JfMmcSPNh94RAj8AAJwKi2OWzcWKzHoKkBDSszfF73hrMQCcCtm6
oYevRaoNA00yrTvL7Qqnk3M=
=P35M
-----END PGP SIGNATURE-----