Wikipedia as Participatory Journalism: Reliable Sources? Metrics for evaluating collaborative media as a news resource | Main | What’s So Special About Wikipedia?

June 30, 2005

Studying Cooperation and Conflict between Authors with history flow Visualizations

Viégas, Fernanda B., Martin Wattenberg, and Kushal Dave. “Studying Cooperation and Conflict between Authors with history flow Visualizations.” CHI 2004. 24-29 April 2004. http://web.media.mit.edu/~fviegas/papers/history_flow.pdf. 30 June 2005.

The authors developed a tool for creating a visual record of multiple document versions in wikis. The goal of this tool is “to make broad trends in revision histories immediately visible, while preserving details for closer examination” (577). Wikipedia served as their research site. The visual maps produced remarkable indications of cooperation and conflict as content was negotiated and vandalism occurred and was fixed.

Results were presented in two versions: with revisions equally spaced, and with revisions spaced according to date. When the space-by-date protocol was used, instances of vandalism virtually disappeared from the record because they were repaired so quickly (often within 2-3 minutes). Five common types of vandalism were identified:

  1. Mass deletion: deletion of all contents on a page
  2. Offensive copy: insertion of vulgarities or slurs
  3. Phony copy: insertion of text unrelated to the page topic
  4. Phony redirection: often pages contain only a redirect link to a more precise term (e.g. “IBM” redirects to “International Business Machines”), but redirects can also be malicious, linking to an unrelated to offensive term.
  5. Idiosyncratic copy: adding text that is related to the topic of the page but which is clearly one-sided, not of general interest or inflammatory; these may be long pieces of text (578-579).

The article also briefly discusses authorship within Wikipedia, noting the usual issues. They discuss familiarity among Wikipedians working on the same pages, and point to inconsistencies between pages regarding anonymous user contribution. (Some have heavy anonymous contribution, some not.) There seems to be no clear connection between anonymity and vandalism. Interestingly, the first version posted of a page seems to remain more intact than sequential edits, a phenomenon the authors term first-mover advantage (580).

Problems were encountered in attempts to measure the stability of Wikipedia pages because of the amount of time needed to develop and run a fine-grained differencing algorithm on all page versions and the fact that, when the research was done in 2003, Wikipedia was very new (581). Researchers therefore “focused on size change as a simple measure of change in content.” Little evidence for stability existed, and many pages with more than 100 versions demonstrated steady growth.

Three possible reasons for Wikipedia’s success were posited as a result of the study:

  1. watchlists, which are unique to Wikipedia, “provide a mechanism for community surveillance,” and may be responsible for the rapid repair of vandlism.
  2. Backchannels (talk pages, list servs) remove meta conversations from the main content.
  3. The Neutral Point of View (NPOV) policy provides “both common ground and rough guidelines” for dispute resolution.
While these reasons seem quite evident now, two years ago they were not. Time has proven them correct.

Posted by kenne329 at June 30, 2005 4:03 PM | Authorship | Community Rules | Conference Proceedings | Wikipedia

Comments