|
 Some years ago, when Yelp was exciting and brand-new, a group of friends — not wanting to take a chance on wasting money at a tourist trap — consulted the novelty app for restaurant reviews of a little Italian restaurant in a huge metropolitan city overrun with little Italian restaurants.
One of the reviews called that particular restaurant a favorite of the notorious gangster John Gotti.
Whether this was a benchmark by which all Italian restaurants are measured, we did not know. Yet, because the Yelp post seemed full of other positives, we went inside.
Clearly, that Yelp review was packed with John Gotti-sized lies because the pricey meal was no better than what you would find at a fast food joint…or worse, at my house.
Fast forward some years later and it turns out that sites similar to Yelp, like TripAdvisor and Amazon, are packed with fake reviews meant to bolster a product or business.
“Misbehaving companies can either try to boost their sales by creating a positive brand image artificially or by generating fake negative reviews about a competitor. The motivation is, of course, money: online reviews are a big business for travel destinations, hotels, service providers and consumer products,” said Mika Juuti of Aalto University.
In fact, some of those reviews seem so legitimate, that people don’t even recognize that they are generated by an algorithm.
Now, researchers from Aalto University led by Juuti have developed an algorithm that can determine whether a review is a fake or a legitimate one.
To begin, researchers trained a machine learning model, a deep neural network, on a dataset of three million genuine restaurant ratings posted on Yelp. Once trained, the model produced fake restaurant reviews character by character using a text sequence of “review rating, restaurant name, city, state, and food tags.”
“In the user study we conducted, we showed participants real reviews written by humans and fake machine-generated reviews and asked them to identify the fakes. Up to 60% of the fake reviews were mistakenly thought to be real,” said Juuti.
From there, the team developed a “classifier” capable of spotting fake reviews. The classifier performed well, especially in scenarios where humans evaluating the legitimacy of a review were stumped.
Do you consult reviews before eating out or making a purchase?
| 
Re: Algorithm Detects Fake Yelp, Trip Advisor Reviews
