Clarkson profs have new way to sort through huge amounts of social media data for cybersecurity and legal cases
Wednesday, June 7, 2017 - 11:48 am

POTSDAM – Two local professors have developed a system to quickly and effectively search vast amounts of social media information to aid cybersecurity concerns and lawyers in civil cases.

Clarkson University’s Assistant Professor of Computer Science Natasha Banerjee and Assistant Professor of Electrical and Computer Engineering Sean Banerjee professors are collaborating with a lawyer to more easily extract activities of interest from dense social media content as part of their research on understanding how humans interact in the real world and in online environments.

The Bannerjees operate Clarkson's Terascale All-sensing Research Studio, or TARS for their research.

As part of that research, they recently developed a way to sort through vast amounts of social media, creating a quick and effective way of locating specific information that can be particularly useful in legal cases, according to a press release from the university.

“People share so much of their everyday life online now,” Banerjee said. “A lot of civil cases depend on this data, so lawyers are looking at social media such as Facebook and Twitter. They search for keywords, but struggle with massive amounts of information to look through.”

Last year, she and her husband were at a wedding when they happened to chat with Michael Simms, a lawyer from West Virginia who was in that exact situation. He mentioned how frustrating it was searching through social media for a case. Sean Banerjee suggested they talk and “it ended up being a productive discussion,” he added. The result is a shortcut through the cyber haystack.

The Banerjees developed an algorithm that automatically scrolls through series of screen shots of social media pages that contain essential phrases and words pertinent to the case. This “image-based text forensics for e-discovery in social media websites” helped Simms settle the civil case he had been working on.

“It was nothing short of amazing. They gave me the information, narrowly tailored to what I needed,” Simms said. “All parties in civil litigation are looking hard at social media. This is a service that others could use, I'm sure. If I need them again, that's who I'm turning to.”

“They managed to provide me with a service that big companies could not provide me with, and in an easy-to-use format,” Simms said. “They really nailed it for me.”

The Banerjees are among the first to provide this type of service and hope to reach out to lawyers and let them know they have the expertise that can help them.

The Banerjees' work also potentially can tie in to Clarkson University's interests in cybersecurity. They submitted a paper about their work to the 2017 Annual Symposium on Information Assurance (ASIA). The paper was accepted as a publication, so they will be presenting it at the ASIA conference that will be held this June in conjunction with the New York State Cyber Security Conference in Albany.