URI_Research_Magazine_Momentum_Spring_2015_Melissa-McCarthy

restaurant owner could use the tool to show all negative reviews came on a Tuesday and were all written by recently registered reviewers. The owner could then tell Yelp that based on URI’s work something seems fishy, and add that the restaurant is closed Tuesdays. Other cases may be more subtle and based on a complex analysis of the review details. “I think this is another way we can help not only the public but also small business owners,” Sun says. “We provide the scientific results and they provide the context.” Sun never set out to research product reviews when she arrived at URI in 2004. She held an interest in trustworthy computing — the concept of ensuring that millions of interconnected devices are secure and not stealing credit card numbers or, say, shutting down the power grid. With so much data today streaming through the air, Sun honed in on securing wireless networks. She started analyzing signals generated from network equipment. She found that by applying her expertise in signal processing, she could identify patterns. Devices that did not follow an expected pattern could be labeled as suspicious and throttled (slowed down or regulated) or blocked. Highly trusted devices could be given priority access. Sun started thinking about how the theory could be applied to other areas in life. Then, one day shopping online for her family, it dawned on her that the concept could help her ferret out fake reviews. “This is another case where social science helps computer engineering and vice versa,” she says. Online reputation management is a booming field. As Americans spend more and more time living and shopping in a virtual world the line between physical and online blurs. Sun notes customers who meet a saleswoman in person but cannot find her company online may question the validity of the business. Prospective college students exploring a high-tech program who find a rudimentary website apply elsewhere. “Online reputation is getting connected to offline reputation,” Sun says. “That’s

Yan Sun Associate Professor

Electr ical , Computer and Biomedical Engineer ing

why our research is so exciting and important.”

± ± ±

Spring | 2015 Page 53