In order to efficiently gather information about preventative measures and strains of a virus during flu season, Professor Dredze of Johns Hopkins was able to screen out useful tweets relating to the flu and use the open CDC (Centers for Disease Control and Prevention) data to create a predictive model that nowcasted flu conditions with a 29 percent increase in accuracy. In 2013, the CDC, having taken notice of this increased interest in social media flu prediction, organized and hosted its first “Predict the Flu Season” challenge. This challenge focused the nine participating teams of researchers on predicting various milestones of the influenza season—e.g., start, duration and peak—which would be the most useful to public health prevention efforts. The model developed by the researchers at Johns Hopkins demonstrates the promise and potential of the effective use of social media data to inform flu prediction. Their influenza model could predict the virus’s future activity one to two weeks before the CDC data alone with the same accuracy.