BLOOMINGTON, Ind. -- The media called it "The Twitter Predictor," and some scoffed at the idea that by analyzing activity on the social media tool one could predict economic markets like the Dow Jones Industrial Average. But today, Indiana University associate professor Johan Bollen's work received a rare form of validation: a United States patent.
Working hand-in-hand with Indiana University's Research and Technology Corp., the professor of informatics at IU's School of Informatics and Computing called receiving the patent license "a quantum leap for us" and a "huge milestone."
IURTC actually sought and received the patent -- bringing IURTC's total number of active U.S. patents to 171 -- and will license use of the invention formally titled "Predicting Economic Trends via Network Communication Mood Tracking" to Bollen's start-up company, Guidewave Consulting. IURTC is a shareholder in Guidewave and holds a revenue-bearing license agreement with Guidewave that allows the university to collect royalties from sales.
Bollen and his Ph.D. student Huina Mao, co-inventor of the mood tracking system that analyzes hundreds of millions of tweets each day, first gained attention after posting a research paper, "Twitter mood predicts the stock market," at the open access online science archive arXiv on Oct. 16, 2010. After two days, Google had returned nearly 70,000 hits on the title, media picked up on the story, and within a few months hedge funds were offering to invest millions of dollars in their system.
But it's today's official issuance of U.S. Patent No. 8,380,607 that confirms Bollen's and Mao's work as unique, novel and worthy of protection.
"It's an important milestone for a start-up to receive a patent, as it is very likely to help make the company successful," said IURTC Director of Technology Transfer Bill Brizzard, who worked with Bollen on the patenting process that took over two years. "Would-be partners who might have been reluctant in the past will now see this as a safer investment."
The patent also means competitors will be excluded from using the same system. Instead, they'll need to buy the information from Guidewave.
"The past two years have seen tremendous growth in this industry, some of it possibly inspired by our work," Bollen said. "So the patent is quite relevant to the success of our business. Our efforts are being acknowledged by its issuance."
The network tracking system calculates indicators of the public mood state along a multitude of dimensions. The original work used six mood categories -- tension, depression, anger, vigor, fatigue and confusion -- but those have since been expanded to provide an even more accurate and complete picture of changing public and economic conditions. By tracking the content in real time of what is now up to 500 million tweets per day, the network system can detect subtle changes in public conditions that are correlated to specific entities like the Dow Jones and various other financial and economic indicators. Bollen describes it as a process that is constantly on the look-out for interesting statistical patterns in social media, finding the proverbial needle in a haystack.
"A lot of companies can ask people how they feel about a specific brand, product or topic, but we're probing the underlying mood state of entire communities," Bollen said. "We're focused on offering our analytics, the data, to a wide variety a domains, from hedge funds and banks to government agencies and even personal investors."
The partnership between IURTC and Guidewave has been a strong one, both Bollen and Brizzard noted, and the patent and licensing process has been one designed to benefit IU and Indiana.
"The purpose of us licensing from IURTC is to give back to our school, the university and the state of Indiana," Bollen said. "We want to make sure that the community benefits from our work."
"We're trying to provide a benefit to the public by making use of the research that happens at IU," Brizzard added.
AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert! system.