UC Santa Barbara's Lars Hufnagel, along with Theo Geisel and Dirk Brockmann from the Max Planck Institute, have taken an Internet curiosity involving the travels of dollar bills and used it to create a breakthrough model of human travel patterns, one that could greatly boost our ability to respond to emerging epidemics.
Back in the days when few people ever traveled more than a few miles from their homes, diseases spread in a fairly linear way, slowly marching across the countryside. Today, the extreme mobility of the average industrialized country citizen (and their general aversion to wearing tags in their ears to facilitate tracking) makes figuring out how travel patterns relate to epidemic spread a difficult task. The research team theorized that tracking currency -- cash money -- would be the next best thing to collars and tags. But where could they find enough data about how individual bills migrate around the country? The Internet, of course.
Where's George? is a website allowing individuals to enter the serial numbers of the dollar bills (and other US currency) and their location, then to return later to see where those dollars have gone. It's an online curiosity -- not really a game, more like an information toy. As of this afternoon, over 76 million bills have been entered into the site, totaling over $430 million. The research team took this enormous wealth of data and found that human domestic travel patterns, as represented by currency, matched an unexpected -- but easily understood (for mathematicians, at least) -- scaling and diffusion model. This model will make it possible to build far more accurate simulations of epidemic spread and response, potentially saving many lives.
Sometimes, the methods used to discover something overshadow an otherwise fairly useful research result; this may be one of those cases, particularly due to the nature of the discovery. The summaries of the article (which appears in the January 26 edition of Nature, so is behind a subscriber wall) quite reasonably focus on the mathematical results. For the curious, they are:
First, the distribution of travelling distances decays as a power law, indicating that trajectories of bank notes are reminiscent of scale-free random walks known as Lévy flights. Second, the probability of remaining in a small, spatially confined region for a time T is dominated by algebraically long tails that attenuate the superdiffusive spread. We show that human travelling behaviour can be described mathematically on many spatiotemporal scales by a two-parameter continuous-time random walk model to a surprising accuracy, and conclude that human travel on geographical scales is an ambivalent and effectively superdiffusive process.
Epidemiologists will thrill to this discovery, but for most of the rest of us, the novelty comes from the use of what amounts to an idle laugh on the web (the creators of Where's George say they created the site "for fun and because it had not been done yet") as the basis for serious -- and potentially quite valuable -- research into how travel relates to the spread of disease. The value of Where's George came from mapping currency transfer to physical movement; there may not be many other studies that will require the same kind of data. But the bigger idea arising from this is that narrow, quirky social networks may turn out to be useful collections of scientific data about human behavior, if we're willing to look beyond the surface amusement.
