After studying Twitter messages for 10 months, a team of scientists from Indiana University (USA) has concluded that messages from the popular microblogging service can be used to predict movements in the Dow Jones stock index .
Specifically, the researchers analyzed the emotional content of 9.7 million messages written by 2.7 million Twitter users , using the Google algorithm ‘Profile of Mood States’ (GPOMS), and took into account six parameters: tranquility, alertness, security, vitality, kindness and happiness.
In their research they found that changes in a specific parameter, tranquility , were correlated with movements in stock prices. So much so that the accuracy of the predictions of the closing values of the Dow Jones based on this parameter was 87.6 percent, with an advance of 2 to 6 days .
How to use Twitter in 8 steps