"Wikipedia Loops" is a project that creatively navigates layers of the massive, crowd-sourced online encyclopedia Wikipedia. Starting with a seed page, a path through the site is created by following the first link in each Wikipedia entry until a page is reached that has already been traversed. The result is surprising: all articles start from the periphery and work their way toward shifting but common centers of meta-knowledge such as Philosophy, Science, and Language. This paper outlines the methodology used in the Wikipedia Loops project and its relationship to historical models mapping human knowledge.
The final version of the "Wikipedia Loops" project is exhibited on a video monitor with loops pulled periodically in real-time from the Wikipedia server. The loop is parsed and displayed for several minutes, making the result much more akin to a drawing or poetic text than an interactive software installation.
Rather than reinforce the ‘sublime data’ model that many utilitarian data visualizations fall under, "Wikipedia Loops" pares down to the individualized, perhaps random but certainly particular pathways through one of the largest repositories of English-language human knowledge. The process then is more like a derivé through a massive digital landscape. Like a derivé, the starting point is in many ways irrelevant (however interesting), the end-point is not predetermined, nor is the route to get there: it is the unexpected connections and chance encounters that are the reason for the exercise.
A longer discussion of this project's methodology and its relationship to the practice of data visualization will be published spring 2013 in the Leonardo Electronic Almanac (MIT Press).