Twitter Mood Predicts The Stock Market
Tweets about emotional states of calm or anxiety can be used to track the ups and downs of the Dow Jones Industrial Average. Amazing!
Professors at the school of Informatics and Computing looked at 9.8 million tweets posted by approximately 2.7 million users from Feb. 28 to Dec. 19, 2008. Professor Johan Bollen evaluated public sentiment as expressed in tweets to see whether that could be used to help predict the stock market. What they found was that the standard prediction models were improved substantially when certain mood dimensions from their data were included.
“We find an accuracy of 87.6% in predicting the daily up and down changes in the closing values of the DJIA and a reduction of the Mean Average Percentage Error by more than 6%.”
That is truly fascinating and goes to show the impact the social networks are having upon us. They literally move the markets! Read the paper at arxiv.org
About Me
I created my first web site in 1994 and have been working full-time designing web sites since 1996. I currently work as an UX specialist and Information Architect.