Sunday, August 16, 2009

There is a lack of research on Wikipedia

For me while interesting, it is hardly new and therefore not that interesting what people like Ed H Chi write about Wikipedia. They do not write about Wikipedia, they write about the English language Wikipedia. Invariably news written about Wikipedia concentrates on just one of over 260 projects. It diminishes what Wikipedia is about and it ignores important things that are happening.

I would be interested in more study looking at the "other" wikipedias. This is where all kinds of other phenomena exist.

Yesterday Siebrand observed that there is a group of languages that have solid localisations and, the current localisation rally makes this group stand out even more.  We have the impression that this coincides with the vitality of projects; German French Dutch are top performers in localisation they have a healthy community and provide a great Wikipedia. For languages like Spanish Turkish Swedish Italian it is still possible for people to take part in the translatewiki.net localisation rally. People who participate on languages like Estonian and Khmer find that they have to concentrate on doing the most used and MediaWiki core messages first (our rationale being that our Wikipedia readers are best served in this way).

With a sample size of 260, it becomes possible to do research into the effect of localisation and the performance of a project. As the LocalisationUpdate extension is being tested for use in the WMF, timely delivery of localisations becomes a reality once it is implemented. This will give the numbers of localisation and performance a much more direct relation with each other... The question is, if someone is interested in the numbers generated by such research..

It is known for languages like Bangla that Wikipedia is the biggest resource in that language in that language, I can imagine that this is true for other languages as well. When a Wikipedia has such a status, it changes the relevance of that Wikipedia for scientists who study thea language. It is interesting to learn what the effects are on the people who use the internet in these languages. With Wikipedia being the biggest resource does this populate the Google search results and, does this make the Internet more of a worthwhile experience?

We know that things like sources, NPOV, BLP are particularly relevant on our biggest projects. On our smaller projects these things do not get the same attention. Here it is more important to have articles in the first place. The make-up of these communities is likely to be utterly different as well. Would it not be nice to understand how our projects are populated and study how it evolves over time? At what stage all kinds of policies start to kick in?

Research, the numbers they provide are important on many levels. They indicate issues, they indicate where we want to put our resources. The lack of research on the other Wikipedias make the other Wikipedias invisible, issues particular to other languages do not get attention and consequently resources needed to address issues are not available.

My argument is that there is a lack of research on Wikipedia, Wikipedia as a whole would benefit from research and indeed where the English Wikipedia's growth is slowing down, there is plenty of room for growth elsewhere of standard encyclopaedic information in the other projects. This in turn will bring up many subjects that en.wp does not cover. The existence of articles on subjects not covered in en.wp are indicative of a bias and once en.wp starts to cover these subjects it will improve its neutral point of view.. Consequently ALL our Wikipedias including en.wp will benefit from research on the "other" Wikipedias.
Thanks,
      GerardM

1 comment:

Purodha Blissenbach said...

I fully agree with your points. One of the things I try to do and very often cannot complete due to lack of time is to write my new articles in the 3 languages that I happen to be able to write well. (4 if you would want to include "simple English") The most interesting thing I observe, when I do write language versions of an article is, that I hardly translate. I rewrite the article an another language, it is similar to one, but is not a translation. Paying a professional translator for instance to do the work, would give closer related results, but not as good articles. I also mostly find every rewrite to be better than the last language version before, thus I have reasons to improve the older ones.
Thank you for listening.