Abstract:
|
Seasonal influenza is a serious public health and societal problem due to its consequences resulting from absenteeism, hospitalizations, and deaths. The overall burden of influenza is captured by the Centers for Disease Control and Prevention's influenza-like illness (ILI) network, which provides invaluable information about the current incidence. ILI is released at an approximately one to two week lag, however, prompting researchers to develop near real-time biosurveillance techniques based on Internet data (e.g., Google or Wikipedia search queries). The potential value added to disease monitoring by Internet data is significant. It is often assumed, however, the value added by Internet data to disease monitoring will extend to disease forecasting. In this work, we explore that assumption by augmenting ILI data with near real-time Wikipedia access log information to probabilistically forecast seasonal influenza in the U.S. We investigate the value added by Wikipedia with respect to predictive performance for recent historical flu seasons.
|