Can Deceptive Reviews Really Be Weeded Out?In reading product reviews, I am sure that we have all read some in which the writer goes on for paragraph after paragraph using adjectives such as ‘fantastic,’ ‘superb,’ or other wording to describe a particular product or service. When that occurs I can’t help but feel not only like the review is untruthful, but that the writer may not really have any experience with the product that he or she is reviewing.

In considering this aspect of reviews, I discovered that the unfortunate reality is that some companies and businesses actually pay reviewers to write glowing reports for their products or services. That means that the reviewer is paid in cash or compensated for their reviews in the form of products by these companies. Unfortunately, since these paid trolls and their rogue reports leave no Web site unscathed, it is hard to know which reviews are legitimate and which are not. To make it worse, these reviews are often spread across multiple Web sites, showing up on any site where the product is offered and the writer can be paid for the review.

To address the situation, students at Cornell University are developing a software that is meant to identify these bogus reviews. Currently, these students state that their software is about 90% effective in finding these bogus reviews whereas, without the software, humans only detected approximately 50% of falsified reviews. In the case of human detection, they also stated that the consumer was more likely to believe the fake reviews of products and services, however, I find it difficult to believe that people are so easily swayed — especially when it comes to spending their hard earned cash.

For myself, I do not believe that any product is 100% perfect, which makes me offer the following suggestions. First, use caution when a review is overly descriptive or if it appears to have been professionally written. Second, I would suggest caution when you read any review that lists only the pluses of the product, omitting any comments on issues that could be improved on. Last, be on your guard if the review compares the product to others on the market in a way that would negatively influence the consumer to avoid the competitor’s product. On the flip side, there are those who find fault no matter how good the product or service is, meaning that we as consumers need to use our best judgement in determining the validity of any review.

So what do I think of the Cornell research? It is definitely interesting, but my feeling is that while using the software may help to filter out some of the bad reviews, it will only be a matter of time before the professional reviewers adjust their writing style in order to slip through the cracks. That being said, in final analysis, is it up to you the consumer to validate, in your mind, if a review is truthful and honest. Just my two cents.

Comments welcome.

The Cornell University report, Finding Deceptive Opinion Spam by Any Stretch of the Imagination, can be found in .pdf format here.