Last year I wrote about the Worldmapper website—a site that helps users visualize the relationships between different data sets and world geography. In the January/February 2007 issue of The Atlantic, P.J. O’Rourke describes one way that Worldmapper might be used to answer a question about the contemporary world—in this case, how one might use the system to predict where the innovators of the 21st century will live.
O’Rourke’s analysis of the tools on the site provides a useful overview of how this site might be used, but also points to a problem with the project. To date, there are only 200 data sets loaded up into the mapping algorithms. If the folks at Worldmapper want to maximize the use of this intriguing product, they need to either make the software open source so that users can begin tweaking it to make it even more usable, or they need to open up the data repository so that users can begin to deposit their own data into that repository for mapping.
If the algorithms were to become open source, I suspect historians would be early and heavy users of the system. Given the popularity of social history over the past several decades, many of us already sit on large data sets that could usefully be loaded into such a system for display and analysis. Moreover, if the algorithms became open source, I suspect you’d see a lot of interest in repurposing them at the local level. Imagine, for instance, a similar map of the United States in which the 50 states expanded or contracted according to their relationship to one another in the particular dataset.
In my own case, I am sitting on a very large database of sociological and political data from the Bohemian lands between 1880-1910. I also have a map of this territory that shows the individual census districts from which all those data were derived. If I had access to the Worldmapper algorithm, I could create a digital version of that map and load in my data. Census tracts in Bohemia or Moravia would expand or contract according to the rate of growth in industrial production or according to the popularity of radical nationalist politicians.
Because I’ve mined those data so hard over the past several years, I’m not sure whether I would come up with new insights into the relationships between things such as industrial growth and the popularity of radical nationalism—but I might. And I know I’d be able to create some very attractive graphics to show my students those relationships on screen.
I’ve written to the folks at the University of Sheffield who are responsible for Worldmapper and I’ll pass along their response here once I hear back from them.