Digital Innovation & Entrepreneurship
AI AND DIGITAL BUSINESS
HOW CAN BUSINESSES COMBAT FRAUDULENT REVIEWS? by Ram Gopal
TO THE CORE
1. Companies like Amazon are struggling to keep up with fraudulent reviews because machine learning is not enough by itself. But human assistance can train algorithms to detect fakes more efficiently and accurately. 2. Research revealed that fraudulent reviewers are more likely to post uniformly extreme views in flurries of activity from short-lived accounts. Fake reviews are likely to be shorter, less detailed, and more predictable. They are more likely to repeat words and contain errors. 3. Encouraging customers and
reviewers to interact can provide more evidence to detect spammers.
4. The same approach could improve the tools that social media platforms use to detect automated accounts that spread fake news.
I t’s a familiar scenario. You’re planning a week away, a meal at a restaurant, or a new purchase. You want to make the right choice, so you check the reviews left by previous customers. But can you really trust the glowing recommendations and horror stories you find online? Several high-profile legal cases suggest not, exposing how some companies use fake reviews to boost their brand image, promote individual products, and tarnish their rivals’ reputations. Chinese tech giant Alibaba sued a third-party seller in 2016 over a ‘brushing operation’, where customers paid for products and submitted positive reviews before recouping their money. And, in 2019, the US Federal
Trade Commission launched a landmark case targeting phoney Amazon reviewers who left five-star recommendations for a weight-loss supplement that caused liver failure. These are only the tip of the iceberg. Many more cases slip under the radar. Companies like Amazon are struggling to keep up with the flood of fraudulent reviews, despite using algorithms to identify fakes. Machine learning is simply not enough by itself. The technology needs a helping hand. We analysed more than 260,000 real-world restaurant reviews collected from Yelp. This covered 5,044 restaurants across four US states over a five-year period. In doing so, we identified 12 new characteristics shared by many fraudulent
Warwick Business School | wbs.ac.uk
wbs.ac.uk | Warwick Business School
38
39
Made with FlippingBook Learn more on our blog