Never have scientists had so much readily accessible, real-time data about what people say. Twitter, the service that allows users to send text updates of up to 140 characters out to the public, publishes more than 200 million messages, or tweets, a day. Compared with information from cellphone records and social-media sites, Twitter texts are as timely as a pulse beat and, taken together, automatically compile the raw material of social history.
As Twitter’s message traffic has grown explosively, so has the scientific appetite for the insights the data can yield. Dozens of new scholarly studies over the past 18 months by computer-network analysts and sociologists have plumbed the public torrents of data made available by Twitter through special links with the company’s computer servers. This research has harnessed the service to monitor political activity and employee morale, track outbreaks of flu and food poisoning, map fluctuations in moods around the world, predict box-office receipts for new movies, and get a jump on changes in the stock market.