Research talk:Monthly wikimedia editor activity dataset
Welcome researchers and curious others.
Something with mobile?
Just a wild guess, but could there be any correlation with improvements in mobile experience? I personally had been "neglecting" Wikipedia for a year or two, but found it more enticing to contribute when I found out that there is a nice Wikipedia app for Android now.
An even much more speculative guess is that the whole evolution since 2001 is somehow tied to the state of the economy. Around 2007, crisis sets in and ostensibly leaves people less room in their lives to edit Wikipedia, around 2015 they ostensibly begin recovering that room?--Anders Feder (talk) 03:12, 26 September 2015 (UTC)
Hobby/Side Project in Pairs/Groups?
Hi - might anyone be interested in co-coding/co-thinking on how to process this data set? IT seems eminently suitable for map/reduce tools. I don't have a lot of familiarity with the latest toolsets for this sort of work but I do have a basic background in stats/data science etc. ----Siruguri (talk) 22:44, 29 September 2015 (UTC)
There were so much fuss over whether to have monthly data open in xtools per default and now we have this. Not like it was impossible to run the queries against replications or API before, just an idle thought. --Base (talk) 22:03, 27 October 2015 (UTC)
- ze Germans disliked the web interface, AFAIK. As for this dataset, it's similar to what provided by Wikistats csv with a few additional columns and rows. The data has been available on the web for... a decade? Nemo 07:11, 28 October 2015 (UTC)
Maybe Wikipedians now have a new activity that makes them edit more. E.g., if we were rebuilding the category system, the result would be a lot of more edits. Or the Wikipedians do something differently than before. Could it be e.g. a repercussion from Wikidata in some indirect way? Is it a bot notifying defunct weblinks, causing humans to repair them? Ziko (talk) 19:07, 30 December 2015 (UTC)
So what happenned in early 2007?
I could understand a plateau after a sharp climb. I could understand a slow decline from the plateau state. I cannot image what happened in early 2007 to turn a yearly growth of +1000 active editors into an almost equally sharp decline of hundreds of editors year-on-year. Did policies change? Did interface change? What changed so rapidly that editors started fleeing the website all of a sudden? My best guess: statistics methods changed somehow so we are looking at broken data. --SSneg (talk) 19:18, 26 February 2016 (UTC)