Around the Net

Researchers Developing Fake Review Detection

  • CNET, Wednesday, July 27, 2011 11:44 AM
Are product- and service-rating services destined to be overrun by fake reviews? Not if Cornell University has anything to say about it. School researchers have developed software they say can detect fake reviews. "The researchers tested the system with reviews of Chicago hotels," reports CNet. They reportedly grouped 400 truthful reviews with 400 deceptive reviews for the study, and then trained their software to spot the difference. The software apparently got it right about 90% of the time.

"This is a big improvement over the average person, who can detect fake reviews only about 50% of the time," CNet writes, citing additional findings by the researchers. People apparently fall into two categories when it comes to judging authenticity. One type accepts too much at face value and doesn't reject enough fake reviews, while the second type is overly skeptical and rejects too many authentic reviews. Despite their different approaches, each camp is right about half the time.



Read the whole story at CNET »

Next story loading loading..