IU professor's software uses tweets to predict Wall Street - 13 WTHR Indianapolis

IU professor's software uses tweets to predict Wall Street

Posted: Updated:
A program developed by an IU professor reads tweets to determine the mood of its users. A program developed by an IU professor reads tweets to determine the mood of its users.
The mood is then translated into a forecast for the stock market. The mood is then translated into a forecast for the stock market.
Associate Professor Johan Bollen developed the software. Associate Professor Johan Bollen developed the software.
BLOOMINGTON, Ind. -

The twittering thumbs of millions of Americans may be just as powerful as corporate dealmakers when it comes to Wall Street.

An IU professor and researcher just received a patent for software that crunches hundreds of millions of tweets, to predict where the stock market is headed.

Random, seemingly frivolous and hurriedly-typed tweets like "egg salad joke" and "There are a lot of pools in Texas" that some, like IU graduate student Meghan Ploch, find annoying, may prove useful.

"People writing about pointless things, like, 'I'm eating a sandwich' or 'sitting in a car,' whatever they are doing in the moment," she said.

The tweets may, in fact, be valuable indicators of where the stock market is going.

"The thing is, on Twitter, people do actually tell you how they feel," explained IU Associate Professor Johan Bollen.

His now-patented software sorts and scans a half-billion tweets a day.

"Picks up the most important topics, reads the tweet and assigns values to them," he said.

After crunching the numbers, the software measures the country's collective mood.

"Angry, calm, anxious, confused. There are a variety of measurements we can use to assess the mood state," Bollen said.

It comes down to the tweeting power of numbers.

Think of this way: The thoughts of two or three million people probably don't add up to much, but if you multiply that by tens or hundreds of millions of people, then you may have something.

"We find that when people get more anxious, then there is a great likelihood of the market dropping 3-4 days later and vice versa," Bollen said.

Bollen says his research found that in simulated tests of the software, financial institutions out-performed the stock market 60 to more than 80 percent of the time.

"Well above the chance level," he said with a smile. "It's really astonishing."

Bollen and his fellow researchers have been working on the software for years. Now that economy is picking up, financial institutions, retailers, and other industries give their invention a thumbs up and put it to work.

Powered by WorldNow