Show Hide image

Twitter can 'predict stock market fluctuations'

Tracking 'mood' on Twitter feeds shows 88% accurate predictions of changes in Dow Jones closing valu

The stream of words that emanate from Twitter can predict the ups and downs of the stock market, new research shows.

Researchers from the University of Indiana-Bloomington and University of Manchester found an accuracy of 87.6 per cent in predictions of the daily up and down changes in the closing values of the Dow Jones Industrial Average by tracking the 'mood' on Twitter feeds three or four days earlier.

When the researchers compared the national mood in the US to the Dow Jones Industrial Average, they found that calmness corresponds well with the fluctuations of the stock market.

The researchers trained a machine-learning algorithm to predict whether the stock market would go up or down, using only the Dow Jones Industrial Average from the past three days. Adding emotional data increased the accuracy of the predictions.

The researchers analysed 9.8 million tweets from 2.7 million tweeters between February and December 2008 and selected those that indicated an emotion -- tweets that included words "I feel" or "I'm feeling," for instance -- and ran the test on a Google database to come to their conclusions.