Submissions/Snuggle: Software support for wiki-mentors

From Wikimania 2013 • Hong Kong
Jump to navigation Jump to search
Yes check.svg

This is an accepted submission for Wikimania 2013.

Video of presentation

Logoyoutube2011favicon.svg Watch session on YouTube

Snuggle logo.png
Submission no. 
Subject no. 
Title of the submission
Snuggle: Software support for wiki-mentors
Type of submission
Author of the submission
Aaron Halfaker
Country of origin
United States of America
University of Minnesota
E-mail address
Personal homepage or blog

There's a declining number of active editors on the English Wikipedia. Recent research exploring the cause of this decline has shown that: (1) the decline in active editors is due to a decline in the retention of good-faith newcomers, (2) declining retention appears to be caused by an increasingly negative environment for newcomers that was exacerbated by the introduction of counter-vandalism tools like Huggle and ClueBot NG[1], and (3) Wikipedians current means of identifying and supporting these desirable, good-faith newcomers can’t keep up[2][3]. This presentation will introduce Snuggle, a software system designed to enable current Wikipedians to more quickly identify and support desirable newcomers in Wikipedia. The presentation will briefly overview the system design, demonstrate the tool's functionality and present a new results describing the effectiveness of the tool in increasing retention of desirable new editors.

Detailed proposal

In 2006, the English Wikipedia faced both an amazing opportunity and what was widely believed to be an insurmountable problem. The open encyclopedia was growing exponentially both in new content and new contributors. Around this time, a now highly cited study was published in Nature that showed the encyclopedia wasn't just being filled with *something*, but that the content was comparable -- and in some ways better -- than traditional encyclopedias[4]. But along with this success and growth came a problem -- anonymous vandalism.

Those unfamiliar with Wikipedia were skeptical of it's success because of its openness. In Wikipedia, the content is contributed openly by internet users anonymously. Along with Wikipedia's openness to new encyclopedia content came openness to vandalism. Intuitively, many assumed that the whole system was poised at the edge of destruction as soon as the vandals outnumbered the good-faith contributors who kept them at bay. As the English Wikipedia gained in popularity, Wikipedia's resilience to damage was tested[5].

In response, Wikipedians turned their rules and tools into a complex immune system based on a mixture of anti-vandalism strategies[6][7]. From robots (initially simplistic, but increasing in complexity and effectiveness) that catch the most egregious vandalism and human computation systems (e.g. en:WP:Huggle) that combine human judgement with computational efficiency to interface improvements that made manually removing damage and warning vandals take less time away from actually improving the wiki, Wikipedians built themselves efficient tools to minimize the amount of effort necessary to keep the wiki clean[8].

Along with efficient vandal fighting came a cost that remained hidden until recently. In the effort to improve the efficiency with which vandals and other undesirable editors were turned away from the Wikipedia, desirable new editors were caught in the crossfire. In early 2007, the English Wikipedia's exponential growth in active editors abruptly changed directions and entered a steady decline[9][10]. Research shows that this decline was primarily due to a substantial drop in the retention of new, good-faith editors. The evidence suggests that the drop in retention was a direct effect to the impersonal environment that anti-vandal tools and the warning templates they distribute[11][12]. In other words, the ingenious technologies that saved Wikipedia from collapse in the face of a wave of vandalism were beginning to choke it to death.

The editor decline. The number of active editors (>=5 edits/month) is plotted over time for the English language Wikipedia.
The Snuggle UI: A screenshot of the Snuggle user interface is shown with newcomers sorted by their "desirability". The best newcomers appear on top.

While Wikipedia had adjusted its strategy to efficiently identify, repair and report the damage caused by vandals, it had failed to build an equivalent socialization system to identify, support and instruct desirable new editors who had come to the encyclopedia with fragile motivation[13][14].

In this presentation, I'll introduce you to, Snuggle, a tool designed to rectify this imbalance. Snuggle (a play on "Huggle", one of Wikipedia's prolific anti-vandal systems) leverages a similar set of strategies to those that made the anti-vandalism systems so effective. Snuggle is a web-based user interface that lists recent newcomers to Wikipedia by a machine generated probability of good-faith activity and allows the user (Snuggler) to either contact the newcomer (support) or label the newcomer as in need of support. It is my intention that this system supplement other initiatives aimed (such as the en:WP:Teahouse) by increasing the efficiency at which Wikipedian mentors can identify desirable newcomers and provide them the support they need to keep contributing productively.

In this presentation, I'll demonstrate the functionality of Snuggle and briefly overview the system's components. I'll also present preliminary results of the effectiveness of Snuggle in improving the rate at which desirable new editors are retained.

  • WikiCulture and Community
Length of presentation/talk
25 Minutes
Language of presentation/talk
Will you attend Wikimania if your submission is not accepted?
If I can afford it, yes, absolutely.
Slides or further information (optional)

See en:Wikipedia:Snuggle.

Special requests


  1. Halfaker, A., Gieger, R. S., Morgan, J., & Riedl, J. (2013). The Rise and Decline of an Open Collaboration System: How Wikipedia's reaction to sudden popularity is causing its decline. American Behavioral Scientist pdf
  2. meta:Research:New_user_help_requests
  4. Giles, J. (2005). Internet encyclopedias go head to head. Nature, 438(7070), 900-901.
  6. Halfaker & Riedl, Bots and Cyborgs: Wikipedia's Immune System, 2011. A massive distributed human cognition system emerged
  7. "Lives of Bots," Critical Point of View: A Wikipedia Reader, G. Lovink ad N Tkacz, eds., Inst. of Network Cultures, 2011, pp. 78-93.
  8. Geiger, R. S., & Ribes, D. (2010). The work of sustaining order in Wikipedia: the banning of a vandal. CSCW (pp. 117-126).
  9. Suh, B., Convertino, G., Chi, E. H., & Pirolli, P. (2009) The singularity is not near: slowing growth of Wikipedia. WikiSym, (pp. 8:1-10).
  10. Editor trends, 2011
  11. Geiger, R. S., Halfaker, A., Pinchuk, M. & Walling S. Defense Mechanism or Socialization Tactic? Improving Wikipedia's Notifications to Rejected Contributors, ICWSM'12.
  12. The Rise and Decline of an Open Collaboration Community: How Wikipedia's reaction to sudden popularity is causing its decline. (2013). American Behavioral Scientist, 57(5), 664-688. 10.1177/0002764212469365.
  13. meta:Research:New_user_help_requests
  14. meta:Research:Newbie_teaching_strategy_trends

Interested attendees

If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with four tildes. (~~~~).

  1. Slashme (talk) 18:22, 7 April 2013 (UTC)
  2. ShenMi MeiRen (talk) 12:39, 9 April 2013 (UTC)
  3. អមីរ ឯ. អហរោណិ 06:52, 12 April 2013 (UTC)
  4. Jdforrester (talk) 11:29, 12 April 2013 (UTC)
  5. Jtmorgan (talk) 23:45, 17 April 2013 (UTC)
  6. Daniel Mietchen (talk) 22:45, 22 April 2013 (UTC)
  7. DarTar (talk) 21:53, 29 April 2013 (UTC)
  8. Steven (WMF) (talk) 04:50, 30 April 2013 (UTC)
  9. Ocaasi (talk) 17:27, 30 April 2013 (UTC)
  10. Waldir (talk) 17:41, 2 May 2013 (UTC)
  11. Vera (talk) 11:30, 5 August 2013 (UTC)
  12. Blue Rasberry (talk) 09:29, 10 August 2013 (UTC)