An article in the Press of Atlantic City notes that Twitter can act as an extremely accurate early warning system for the spread of disease.
Adam Sadilek at the University of Rochester analyzed 4.4 million tweets from users in New York City and was able to predict when healthy people would get sick up to eight days in advance, with about 90 percent accuracy.
Given that three of your friends have flu-like symptoms, and that you have recently met eight people, possibly strangers, who complained about having runny noses and headaches, what is the probability that you will soon become ill as well? Our models enable you to see the spread of infectious diseases, such as flu, throughout a real-life population observed through online social media.
Sadllek applied machine learning and natural language understanding techniques to determine the health state of Twitter users. Focusing on tweets that were geo-tagged his team was able to plot them on a map and observe how sick and healthy people interact. Using this data Sadilek is able to predict if and when an individual will fall ill with a high degree of accuracy and potentially predict the emergence of global epidemics.
The video below is heatmap visualization of flu in New York City, as observed through public Twitter data.
Previous methods (including Google Flu Trends and government data) entail time lags from days to years, whereas Sadilek use of Twitter data allows him to show emergent aggregate patterns in real-time, with second-by-second resolution.
Sadilek also developed a web application called Fount.in that monitors the health of New Yorkers based on their Tweets.
These fine-grained epidemiological models could even power a location-aware iPhone app that warns you when you are in an area with a high incidence of flu so that you can avoid it or take the necessary precautions.
There isn't an app for that. Yet.