In the last blog post I wrote about the importance of reviewing and rating content. I briefly mentioned all the problems that Amazon had been having with fake reviews. This problem caused Amazon to receive a lot of bad press, and one wonders if the result of this bad press cut into their bottom line. Authors write their own fake reviews, their friends and family write fake reviews, and some pay others to write fake reviews.
The magnitude of fake reviews on the internet almost defies believability. A New York Times article from August 25, 2012 tells about the people who write fake reviews and the people who buy them. One author spent $20,000 on review services.
A Guardian article from January 25, 2013 claims that fake reviewing is taking place “on an almost industrial scale, with companies paying offshore contractors to post [fake reviews].” The article goes on to explain the details of the world-wide fake review industry, which can be found on websites like Freelancer.com.
Gartner, an information technology research company, came out with a research report in September of 2012, that claims by 2014, 10-15% of all online reviews will be fraudulent. 15% equals 1 in 7 of all online reviews. And then there is the problem of fake Facebook Likes, which I am not even going to get into in this blog post.
So, what’s being done about all this deception and lying? Companies like Amazon, Hilton, and Trip Advisor are hiring researchers who have developed software to spot fake reviews. A team from Cornell has come up with an algorithm that can spot fake reviews 90% of the time (New York Times article 8/20/2011). Software engineers at the University of Illinois have developed a program called GS Rank, or Group Spam Rank, that targets groups who write fake reviews. And yet another approach has been developed by a team of researchers from the Stage University of New York at Stony Brook. You can read about their work here.
On Yelp’s blog page, the company lays out their anti-fake review strategy, which uses shaming sting operations, and automated software called the “review filter.” TripAdvisor also uses software to spot fake reviews and then marks the guilty businesses with a Red Badge that warns users that the reviews may not be genuine (see blog.reviewinc.com). Algarvedailynews.com reports that about 13,000 of Trip Advior’s reviews “fall into a questionable category every day.” The Advertising Standards Authority ordered TripAdvisor to remove the slogan “reviews you can trust”, because so many of their reviews are fake.
With all these problems from user generated reviews, you would think that the best solution would be to avoid them all together. For some websites this is obviously not an option, but with other websites, where customer reviews do not make up the core value of their service, using other rating methods may be more accurate and less controversial. I will talk about some of these methods in future blog posts. Perhaps fake reviews can be eliminated some day, but for now and in the foreseeable future, they remain a huge problem and a constant challenge for websites that rely heavily on customer reviewing.