Warncke-Wang, Morten2017-03-142017-03-142016-12https://hdl.handle.net/11299/185112University of Minnesota Ph.D. dissertation. December 2016. Major: Computer Science. Advisors: Loren Terveen, Brent Hecht. 1 computer file (PDF); x, 148 pages.Over the past 25 years, commons-based peer production [Ben02] has become a vital part of the information technology landscape. There are successful projects in different areas such as open source software (e.g. Apache and Firefox), encyclopedias (e.g. Wikipedia), and map data (e.g. OpenStreetMap). A common theme in all these communities is that they are mainly volunteer-driven and that contributors are able to self-select what they want to work on. Studies on contributor motivation in peer production have found “fun” and “appropriate challenges” to be strong factors [Nov07; LW05], both associated with the sensation of vital engagement often referred to as “flow” [NC03]. Peer production contributors also often refer to altruism, the desire to be helpful to others, as a motivating factor [BH13]. To what extent does this bottom-up, interest-driven, volunteer-based content production paradigm meet the needs of consumers of this content? This thesis presents our work on improving our understanding of how peer production communities produce quality content and whether said quality content is produced in areas where there is demand for it. We study this from three perspectives and make contributions as follows: we investigate what textual features describe content quality in Wikipedia and develop a high-performance prediction model solely based on features contributors can easily improve (so called “actionable features”); we apply a coherent framework for describing and evaluating quality improvement projects in order to discover factors associated with the success and failure of these types of projects; we introduce an analytical framework that allows is to identify the misalignment between supply of and demand for quality content in peer production communities and measure the impact this has on a community’s audience. The research presented in this thesis provides us with a deeper understanding of quality content in peer production communities. These communities have created software, encyclopedic content, and maps that in many ways improve our everyday lives as well as those of millions of others. At the same time we have identified areas where there is a shortage of quality content and discussed future work that can help reduce this problem. This thesis thus lays the foundation upon which we can build improved communities and positively impact a large part of the world’s population.enHuman-Computer InteractionMachine LearningOpenStreetMapWikipediaThesis or Dissertation