[OSM-talk] View data

Morten Wang nettrom at gmail.com
Mon Mar 10 23:47:31 UTC 2014


I logged on to IRC last week and got some good comments there too, so once
again, thanks for asking good questions and providing good help!

One thing that was brought up is the question of how this research will
benefit OSM, how will it make the map better?

We're aiming to discover to what extent areas of the map are popular
(viewed often) but appear to lack quality. These areas are perhaps the ones
the community would be most interested in improving, an argument that was
also put forward in a mailing list post from 2007 that I found (
http://thread.gmane.org/gmane.comp.gis.openstreetmap/17944), although in
that case ways of getting at the quality of a object/area was not discussed
(and most of the URLs appear to return 404s now). While one can argue that
a contributor following their personal interest should suffice, previous
research has suggested it can lead to bias in content coverage. For
instance Haklay's research on OSM suggested socio-cultural bias in the
coverage of London. In Wikipedia there's shorter coverage of films with
primarily a female audience compared to those with a primarily male
audience (Lam et al., WP:Clubhouse…, WikiSym 2011,
http://grouplens.org/system/files/wp-gender-wikisym2011.pdf)

Understanding more about what these parts of the map are can then
potentially lead to development of software tools that makes this
information more readily accessible for those members of the community that
are interested in helping out with that.

By the way, the research papers that we publish are now usually also freely
available on our website (see http://www.grouplens.org/biblio), I saw lac
of access was a mentioned in some mailing list posts. When it comes to
studying something similar to OSM, the research group I'm a member of did
develop a geographical wiki with a bike map of the twin cities of
Minneapolis & St.Paul, see http://cyclopath.org/


Regards,
Morten



On 4 March 2014 17:27, Morten Wang <nettrom at gmail.com> wrote:

> Thanks for your replies, Fredrik and Andy!
>
> I'll get on IRC tomorrow and see if I can get in touch with someone there.
>  I'm aware that we might be asking for something that isn't possible, so
> I'm only cautiously optimistic.  The usage of CDNs is a good point I hadn't
> thought of, I've added a note about that, thanks!
>
> Andy brings up a valid concern, something we've also thought about here.
>  We'll of course have to take that into consideration in any analysis we do.
>
> Thanks again!
>
>
> Cheers,
> Morten
>
>
>
> On 4 March 2014 05:44, SomeoneElse <lists at mail.atownsend.org.uk> wrote:
>
>> Morten Wang wrote:
>>
>>> Hi,
>>>
>>> One of my current research projects[1] looks at OpenStreetMap and we're
>>> interested in knowing the number of views for different regions in North
>>> America (with North America as defined by Geofabrik[2]).
>>>
>>
>> and (apologies for stating what might be obvious but) it's difficult or
>> impossible to get all views of "the data" - any logs from osm.org will
>> miss anyone accessing the data from any other source - be it standalone
>> maps on phones and other devices, maps rendered by other people, use of the
>> data other than in maps, etc. etc.
>>
>> Cheers,
>>
>> Andy
>>
>>
>>
>> _______________________________________________
>> talk mailing list
>> talk at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20140310/e937f640/attachment.html>


More information about the talk mailing list