How monitoring how Twitter users are feeling can predict the stock market three days in advance
Posted by Xeno on October 21, 2010
– The general mood among Twitter users can predict the rise and fall of the stock market almost a week in advance, a new study has found.
A computer scientist in the US has discovered that the correlation between the Dow Jones and the collective public mood was almost 90 per cent accurate.
The link raises the startling possibility that stockbrokers may one day be able to make bets on the stock market based purely on how people are feeling on Twitter and blogs
The remarkable discovery came after researchers at Indiana University noticed the correlation after they analysed more than 9.8 million tweets from 2.7 million users during 10 months in 2008.
The team picked tweets which described how the user was feeling and then these descriptions were divided into six categories of emotion: calm, alert, sure, vital, kind, and happy.
An online mood-tracking tool called OpinionFinder also measured whether the users were generally positive or negative and how that changed over time.
The team was then able to measure variations in public mood and then compare them to closing stock market values. …