Media Matching
Identifying barriers faced by experienced editors in English, Arabic, and Japanese Wikipedias when adding media to Wikipedia Articles.
Overview
The Media Matching project aimed to uncover barriers faced by experienced editors when adding images and other media to Wikipedia Articles. This research was conducted to help inform a grant-supported three year development plan that aims—in part—to increase the number of images used across different language versions of Wikipedia.
Conducted via semi-structured interviews with editors recruited from English, Japanese, and Arabic Wikipedias, this effort revealed that experienced editors in all three communities in fact face different sets of barriers. While English editors note weaknesses in the tools available to them while working with media on Commons, Japanese editors note a culture of sensitivity to image license and copyright on Japanese Wikipedia that raises the bar for images’ inclusion. Arabic editors, on the other hand note a lack of support for the Arabic language on Commons which makes it more difficult to use images that might otherwise be Wikipedia-appropriate. Moving forward, non-English editors may benefit from better integration of machine translation into Commons and Wikipedia at various points in order to allow them to better interpret and use material taken from Commons.
Top High Level Takeaways:
- Making the jump to Commons: Non-English editors often experience a “hard break” between Wikipedia and Commons. Making the jump to the primarily English Commons from their home Wikipedias can be difficult.
- Finding media on Commons: Experienced English editors tend to navigate Commons using categories above all else. Non-English editors typically find themselves relying on machine translation at some point before they access Commons, because searching on Commons tends to be significantly less functional in non-English languages.
- Captions: English editors often copy image descriptions directly from Commons. This is more laborious for non-English editors, who must often either rely on machine translation or compose a new caption, because Commons file descriptions are not appropriate in most non-English Wikipedias.
- Localizing maps and diagrams: The Arabic community in particular places great emphasis on translating images such as maps and scientific diagrams into Arabic. However, the tolls available to support this time-consuming and skilled work remain limited. In turn, the supply of such images that are appropriate and useful for the Arabic Wikipedia readership is also limited.