Work-to-Rule: The Emergence of Algorithmic Governance in Wikipedia
0 likes1,107 views
The document discusses the emergence of algorithmic governance in Wikipedia, highlighting how automated systems and bots play a central role in managing content and rules. It notes that the shift towards algorithmic processes has made rules less transparent and harder to change while proposing that existing cultural differences should guide the adaptation of these governance mechanisms. Recommendations include investigating how rules can be integrated into technical infrastructure and designing features to enhance the visibility of algorithms for non-programmers.
Work-to-Rule: The Emergence of Algorithmic Governance in Wikipedia
- 1. Claudia Müller-Birn1• Leonhard Dobusch1 • James D. Herbsleb2 1 Freie Universität Berlin, 2 Carnegie Mellon University International Conference on Communities and Technologies 2013 July 2nd 2013, Munich (Germany) Work-to-Rule: The Emergence of Algorithmic Governance in Wikipedia
- 2. Picture: http://commons.wikimedia.org/wiki/File:Wikimedia_Foundation_Sue_Gardner_Sept_2010.jpg Quote: http://soc.li/uhtOs4Y Data: http://meta.wikimedia.org/wiki/Template:Wikimedia_Growth January 15th, 2013: 12th anniversary of Wikipedia “The more eyes on an article, the better it is. That is the fundamental premise of Wikipedia, and it explains why Wikipedia works.” “More than 1.5 million people in almost every country have contributed to Wikipedia’s 23 million articles.”
- 3. Aniket Kittur, Bongwon Suh, Bryan A. Pendleton, and Ed H. Chi. 2007. He says, she says: conflict and coordination in Wikipedia. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '07). ACM, New York, NY, USA, 453-462.
- 4. Brian Butler, Elisabeth Joyce, and Jacqueline Pike. 2008. Don't look now, but we've created a bureaucracy: the nature and roles of policies and rules in wikipedia. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '08). ACM, New York, NY, USA, 1101-1110. “the “policyless” ideal that wikis represent is a pipedream”
- 5. Cl. Müller-Birn (CC BY-SA 3.0)
- 6. Flagged Revisions Extension « not registered » Cl. Müller-Birn (CC BY-SA 3.0)
- 7. Flagged Revisions Extension • Not registered editors can only indirectly check their contributions. • The semi-manual system of detecting malicious contributions has been replaced by an automated workflow. • Positive faith has been replaced by a negative one. Cl. Müller-Birn (CC BY-SA 3.0)
- 8. German Wikipedia Cl. Müller-Birn (CC BY-SA 3.0)
- 9. German Wikipedia Cl. Müller-Birn (CC BY-SA 3.0)
- 10. Flagged Revisions Extension • Bots are autonomous software programs that make edits based on predefined rules • Bots help their operators to carry out “mindless, boring and often reoccurring tasks” • Bots belong to a special user group with special user rights • Bots should avoid editing activities outside the article namespace Cl. Müller-Birn (CC BY-SA 3.0)
- 11. Flagged Revisions Extension 0 1,000,000 2,000,000 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 boteditcountperyear Category Main Other Portal Talk User User talk Wikipedia editing activities includes additional namespaces editing activities extends to community space focus expands further to project spaces about 45 % of all edits are outside of article namespace Cl. Müller-Birn (CC BY-SA 3.0)
- 12. check syntax edit language add template update template add data update data archive article check for spam set interwiki-link create content list link disambiguation inter-bot-cooperation create to-do list send to-do alert compile statistic enforce rule welcome user support election Cl. Müller-Birn (CC BY-SA 3.0)
- 13. • Content-oriented focus • Almost 97% of all analyzed bots take over tasks in this area check syntax edit language add template update template add data update data archive article check for spam set interwiki-link create content list link disambiguation inter-bot-cooperation create to-do list send to-do alert compile statistic enforce rule welcome user support election Cl. Müller-Birn Attribution-ShareAlike 3.0 Unported (CC BY-SA 3.0)
- 14. check syntax edit language add template update template add data update data archive article check for spam set interwiki-link create content list link disambiguation inter-bot-cooperation create to-do list send to-do alert compile statistic enforce rule welcome user support election • Task-oriented focus • 6% of all analyzed bots take over tasks in this area Cl. Müller-Birn (CC BY-SA 3.0)
- 15. check syntax edit language add template update template add data update data archive article check for spam set interwiki-link create content list link disambiguation inter-bot-cooperation create to-do list send to-do alert compile statistic enforce rule welcome user support election • Community-oriented focus • 1% of all analyzed bots take over tasks in this area Cl. Müller-Birn (CC BY-SA 3.0)
- 16. What is the guideline about? “Hi, let’s talk about something different - expired votes: How should we deal with them? My vote has been crossed out and I do not agree with this.” • An administrator is automatically re-elected if users with voting rights support this re-election request. • Either 25 votes within one month or 50 within six months are needed. • Votes are only valid if they are submitted (by signing a page) in one of the two possible periods of time . Cl. Müller-Birn (CC BY-SA 3.0)
- 17. “This would be a nice job for a bot.” “I created a bot request and ask for feasibility” • The final community consensus was to delete these obsolete votes and one bot operator agreed to incorporate this task into her software. • The requirements of how to implement the bot have not been discussed - the bot operator just “extended” an existing program. • A rule is algorithmically translated but its realization is invisible because the source code is not publicly available. Cl. Müller-Birn (CC BY-SA 3.0)
- 18. 2009 2010 2011 2012 2013 Community 0 85 Content 0 7,200 Task 0 340 xqbot’s extending activity focus over time Cl. Müller-Birn (CC BY-SA 3.0)
- 19. • Software features and bots make guidelines and rules part of the infrastructure and to a certain extent, makes them harder to change and easier to enforce. • The conversion of socially developed rules into algorithmic ones makes these rules less transparent. • Algorithmically translated rules are more scalable but they do not handle exceptions well. • The extent and kind of algorithmic governance depends on the cultural background of a community. Let’s summaries Cl. Müller-Birn (CC BY-SA 3.0)
- 20. How can we proceed with these insights? • We might elaborate more on what rules and guidelines can and should be translated into the technical infrastructure in order to help the community to make this decision. • We might investigate existing cultural differences in adapting algorithmic governance mechanisms. • We might design new features that make the algorithms in the source code visible and understandable by non- programmers. Cl. Müller-Birn (CC BY-SA 3.0)
- 21. Aaron Halfaker, R. Stuart Gieger, Jonathan Morgan & John Riedl. (2013). The Rise and Decline of an Open Collaboration System: How Wikipedia's reaction to sudden popularity is causing its decline. American Behavioral Scientist, 57(5), 664-688. One example for stepping in the right direction
- 22. "http://www.flickr.com/photos/racatumba/93569705/ Thank you for your attention. Questions?