Volunteered geographic information (VGI) refers to the geographic subset of online user-generated content. Through Geobrowsers and online mapping services, which use geovisualization and Web technologies to share and produce VGI, a global digital commons of geographic information has emerged. A notable example is Wikipedia, an online collaborative encyclopedia where anyone can edit articles, including those about place. Wikipedia’s editorial transparency and integration with online mapping services make it well suited for studying VGI production.

My dissertation contributes empirical evidence and quantitative methods to an emerging area of study—the efficacy and use of VGI—and I focus on spatial behavior in VGI production, which is largely unknown. In particular, the capacity of a ubiquitous Internet to reduce communication costs has raised questions of whether geographic distance matters in information and economic production. My research tests whether proximity matters in VGI production. That is, do VGI contributors write about nearby places? Moreover, what are suitable research methods to study large-scale VGI system with millions of contributors like Wikipedia?

For my study, I collect a corpus of 32 million contributions to 1 million geotagged Wikipedia articles over 7 years (2001–2008). I use data mining and IP geolocation methods on the corpus to select a sample dataset that includes 7.3 million contributions by 2.8 million anonymous contributors to 0.4 million geotagged articles in 21 languages. To measure the proximity effect between articles and contributors, I develop a “signature distance” metric, which is a weighted average of distances between author and article. To model spatial interaction behaviors of contributors, I use a probabilistic invariant gravity model with an exponential distance decay function. My primary findings indicate that anonymous contributors write about nearby places, and that the influence of proximity decays exponentially and varies categorically.

Keywords. Distance decay; Geotagging; User-generated content; Volunteered geographic information; Wikipedia.