Measuring self-focus bias in community-maintained knowledge repositories

Self-focus is a novel way of understanding a type of bias in community-maintained Web 2.0 graph structures. It goes beyond previous measures of topical coverage bias by encapsulating both node- and edge-hosted biases in a single holistic measure of an entire community-maintained graph. We outline two methods to quantify self-focus, one of which is very computationally inexpensive, and present empirical evidence for the existence of self-focus using a "hyperlingual" approach that examines 15 different language editions of Wikipedia. We suggest applications of our methods and discuss the risks of ignoring self-focus bias in technological applications. Categories and Subject Descriptors H.5.3 [Information Systems]: Group and Organization Interfaces

• Type: Conference
• Year: 2009
• Where: CANDT
• Authors: Brent Hecht, Darren Gergle
• Keywords: Bias, CANDT 2009, Emerging Technology, Topical Coverage Bias, Web 2.0