Hi Han-Tseng, if you or any researchers have ideas as to how to extend glamourous or any of Magnus's tools it is worth asking him, If it sounds interesting to code he might well take the challenge.

I have no idea what "fair" would look like in E coverage, but "unfair" is sometimes easily spotted. My own interests are strongly skewed towards archaeology, so I would tend to cover Egypt or Iraq very differently from someone interested in football or the military. A map of the world's stained glass windows according to coverage on Wikimedia Commons would come up with  huge skew towards England.

Slightly more surprisingly if anyone were to check the home range  of species using Wikimedia Commons they would discover that Llamas and Alpacas also are obviously English - http://commons.wikimedia.org/wiki/Category:Llamas_and_alpacas_in_England being so dominant within http://commons.wikimedia.org/wiki/Category:Lama Though there is an alternative theory that volunteer photographers are biased towards that which is unusual for them.

Things get rather more serious, contentious and sometimes down right murky when you start looking at the cultural patrimony of areas that have had a bloody even genocidal past. There are parts of the world where the record even a century ago shows a different local culture than that which now prevails. I'm currently in the Caucasus where there are several cultures that each tend to take their natural cultural area from whatever their nation's maximum historical borders were, and that's before arguments as to which iron age or bronze age culture is the ancestor of which modern people.

But I agree that we could and should identify gaps and try to fill them, this is something that I hope to contribute to in my new role as wikimedia UK's GLAM organiser. The UK has a huge legacy of global culture acquired during our days of empire, and it seems to me that the least we can do is make sure that digital info is made available to the source countries and cultures.

As for how this could correlate to research opportunities on Commons,  I can see some interesting ones, for starters how long does it take better quality newer images on commons to supplant older lower resolution images in their usage in Wikipedia articles and elsewhere?

Then of course there is the usual Commons minefield around depictions of human nudity. A tiny proportion of Commons images that have generated a very high proportion of the meta level discussion within  the movement, especially betweenCommons and its detractors. I think that would make for an interesting study into cultural conflict on the web.

Regards

Jonathan

On 30 April 2013 04:14, Han-Teng Liao <hanteng@gmail.com> wrote:
Thanks for sharing the great tool k https://toolserver.org/~magnus/glamorous.php!   I am also interested in the geo-linguistic distribution and diffusion of online content. Thus, while I agree with your assessment that legal barrier and digital divide distribution may "distort" the geographic distribution, I am not sure what counts as "fair" distribution. UK has been having a huge online presence and internet infrastructure advantage for various reasons. It is difficult to have a fair distributive "justice" based on counting land masses. Land masses can be a source of distortion as well. ;-)  I would rather use this as "distorted" outcome to advocate for more commons-friendly practices from cultural institutions, particularly from East Asia. 

Also, is it difficult to expand your current tool to generate stats on the language descriptions available? A simple research idea is to see how languages play a role in importing/exporting Wikicommons materials. I used toolserver for my previous research on interlanguage links before. Thus, any collaboration on the toolserver (as opposed to making it available online) might be an option. 

(I reply this email as private email first, and if you find the questions and/or your replies to them are relevant to the mailing list, feel free to reply it to the list as well)

Best,
han-teng liao



On Tue, Apr 30, 2013 at 4:37 AM, WereSpielChequers <werespielchequers@gmail.com> wrote:
If there has been any research it is likely to be out of date. Wikimedia Commons is probably our fastest growing major project - it has grown by a fifth in just the last 8 months. Last August it had only 14 million files and has gained 2.8 million since then.

The contents are much more skewed to mass imports from elsewhere than Wikipedia is, and as those mass imports happen so they change the average composition of the wiki.

Another big influence is national legislation, we probably have far more images from countries which have freedom of panorama laws that are compatible with commons than from countries that don't have such laws.Another factor is historical, countries that have systematically replaced old monuments may have less to take pictures of. Hence Commons has a massive skew towards the UK which has 0.1% of the world's landmass, 1% of its population but rather more than 10% of Commons files. I've uploaded quite a few UK images myself, and I have hundreds of images of Georgia that I can't upload to Commons but can upload to other image sharing sites.

If you are considering research by useage you might find it useful to check https://toolserver.org/~magnus/glamorous.php

Jonathan



On 29 April 2013 19:59, emijrp <emijrp@gmail.com> wrote:
It is very interesting but very little has been researched http://wikipapers.referata.com/wiki/Wikimedia_Commons

The same with other sister projects (Wiktionary, wikibooks, wikisource...).

Research is highly biased towards Wikipedia. Neither other wikifarms have been studied.


2013/4/29 Heather Ford <hfordsa@gmail.com>
Does anyone know of research on images on Wikipedia/Wikimedia Commons?

Thanks in advance!

Best,
heather.

Heather Ford 
Oxford Internet Institute Doctoral Programme 
@hfordsa on Twitter

_______________________________________________
Wiki-research-l mailing list
Wiki-research-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l



_______________________________________________
Wiki-research-l mailing list
Wiki-research-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l



_______________________________________________
Wiki-research-l mailing list
Wiki-research-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l