We have already discussed Google’s ability to predict the spread of epidemics as the flu. Simply by measuring the surge in search by different locations – Google can trace and predict future occurrences and their spread.
As Twitter platform evolves fully on real-time and enjoys a high frequency of reporting, when sudden events as earthquakes occur, people use Twitter to spread news related to the earthquake. That can enable detection of earthquake occurrence promptly, simply by observing the tweets.
A recent academic paper introduced by Takeshi Sakaki, Makoto Okazaki and Yutaka Matsuo from the University of Tokyo investigates the real-time interaction of events such as earthquakes in Twitter and proposes an algorithm to monitor tweets and to detect a target event.
To detect a target event, the researchers have devised a classifier of tweets based on features such as the keywords in a tweet, the number of words, and their context. Subsequently, the have produced a probabilistic spatiotemporal model for the target event that can find the center and the trajectory of the event location.
What they actually did was to consider each Twitter user as a sensor and apply filtering features for location estimation in ubiquitous/pervasive computing. They used both a temporal model (assuming the time of the tweet) and a spatial model (taking into account the location of the tweet). Semantic analyses were applied to tweets to classify them into a positive and a negative class (to distinguish between relevant and irrelevant event descriptions).
As a pilot application, the researchers constructed an earthquake reporting system in Japan.
Because of the numerous earthquakes and the large number of Twitter users throughout the country, they were able to detect an earthquake with high probability (96% of earthquakes of Japan Meteorological Agency (JMA) seismic intensity scale 3 or more are detected) merely by monitoring tweets. The system that was developed to detect earthquakes promptly sends e-mails to registered users. Notification is delivered much faster than the announcements that are broadcast by the JMA.
Using the same Twitter based system the researchers can trace typhoons as well and are planned to track other events as rainbows.
Further research on Twitter as a prediction platform:
Twittter can predict the elections (UK)
Twitter can predict influenza rates (US)