NYC-based startup Verificient Technologies―an alum of the first Kaplan EdTech Accelerator class―was profiled on the Cisco-sponsored blog, "The Network," for its use of "big data" to thwart online course cheating:
As online schools and courses become more popular, there remains a thorny question: How do you, say, make sure the people who take an exam really are the students they purport to be? Or, at a more traditional school, in a room packed with test-takers monitored by a remote proctor, how do you stop cheating?
About three –and-a-half years ago, Tim Dutta, a management consultant and serial entrepreneur, and Rajnish Kumar, a researcher at Georgia Tech, who had worked on a computer visioning and video analytics project for the TSA, came up with a plan to address the problem. Relying on facial recognition, real-time big data analysis and other technologies, they would create a way to monitor cheating without the use of a human proctor. Trouble was, they were a bit ahead of the technology available at the time.
Then, earlier this year, they founded a company, Verificient Technologies in New York City and, last spring, were accepted into Kaplan EdTech's first class. The basic idea works like this: Each school specifies certain parameters to be monitored—say, whether someone gets up from a chair or looks at a cell phone. Then, when students log into the system, it conducts a facial and knuckle scan to verify their identity (finger-printing is too expensive, according to Dutta). As soon as students start taking the test, the system swings into action, videotaping them and monitoring their computers. Then, through an algorithm allowing for real-time video analytics, taking 30 impressions per second, it provides a report flagging possible cheaters in red and suspicious activity in yellow. Officials can click on any questionable section and see a recording of what happened.
READ the entire article "How Big Data and Mobile Technology Startups Are Changing Education," by Anne Field (Oct. 6, 2013) on The Network.
RELATED: Verificient Technologies' Demo Day presentation