Abstract
Wikipedia usage data has been harnessed to estimate the prevalence of influenza-like illness (ILI) in the US population. By observing the number of times certain key Wikipedia articles are viewed each day, a model was developed that accurately estimated ILI, within 0.27% of official Centers for Disease Control and Prevention data. Additionally, this method was able to accurately determine the week in which ILI peaked 17% more often than Google Flu Trends. This work demonstrates the power of open, freely available data to aid in disease surveillance.