Social media research helps identify rumor mills in crises


Social media’s size and speed has evolved the channel into news headlines for many users – especially during times of crises. But its lack of safeguards and journalistic standards by individuals can wreak havoc on accuracy and rumors.

So when a crisis occurs just how can one stem the cycle of posting and re-posting false information? Assistant Professor Yasuaki Sakamoto set out to find the answer after a 9.0 magnitude earthquake hit northeastern Japan in March 2011. Sakamoto, a Japan native, knew first-hand about the fears and concerns that individuals experience when trying to reach family members during a crisis.

As he followed Twitter posts about the earthquake,he saw posts with inaccurate information thatcontinued to be reposted. And false information could spread panic.

With his own experience a driving force, he and post doc Yuko Tanaka together with Ph.D.students Rongjuan Chen and Huaye Li researched the topic and turned the findings into two papers: Perspective Matters: Sharing of Crisis Information in Social Media and Toward a Social-Technological System that Inactivates False Rumors through the Critical Thinking of Crowds. Both papers were nominated for a Best Paper Award at the Hawaii International Conference on System Sciences (HICSS), and one won the award. In its 46th year, HICSSis one of the most prestigious international conferences for Information Systems and Information Technology in the world.

Sakamoto, who joined Stevens in 2006, works in the Center for Decision Technologies.

Prior to the earthquake he was studying how social media spread news about Hollywood stars and new technologies. With a PhD in psychology “I am used to studying how people made decisions and in social media I have tried to develop computational models on how people make decisions,” he says. “But before the earthquake I wasn’t thinking o applications in crisis management and disaster response.”

As part of the research he and his team studied more than 2 million Tweets over a six-month timeline to source the origins of false posts. “Yuko Tanaka’s research has shown that exposing people to tweets that question or criticize accuracy of the false tweets may be able to significantly reduce the spreading of false tweets. The exposure to the ‘criticism’ tweets changes people’s feelings and their intention to spread the questionable tweets.”

The research led by his Ph.D. student Rongjuan Chen identified variables such as geographical proximity to the disaster, social distance to the disaster and feelings – all of which can influence the type of message that’s written. The variables can also serve as a barometer for sharing because different feelings can be an indicator of the likelihood of a post being spread,” says Sakamoto.

Sakamoto is currently in the early stages of analyzing media conversations around Super Storm Sandy. “Comparing Sandy and the earthquake will give us useful information to generalize of our findings.”

The Japan earthquake research papers led to a National Science Foundation grant that he received in 2011. The money has been going toward the development of a Web site where others could document the accuracy of other crisis-related posts. “We would like to have it wrapped up by the summer,” he says.