Amazon is using machine learning to identify the most helpful reviews

Finding quality Amazon reviews has never been as simple a process as many people would like, but with the addition of machine learning software, the company hopes to change this. Amazon’s current review system requires direct feedback from users to get a general idea of which reviews people liked the most, whereas this news system will actually be able to learn which reviews are the most helpful and make it easier for users to see these reviews. 

Amazon is overhauling its reviews system with machine learning software to help display more helpful insights on its products from buyers, reports CNET. The company is rolling out its upgraded system in the US and hopes that it will learn which reviews are most helpful to users and improve its capabilities over time, said an Amazon spokeswoman. Newer reviews, as well as those from verified Amazon customers and those reviews that other users vote up for being helpful, will gain prominence on product pages. Similarly, product ratings, which were previously only an average of all review scores, will also become weighted using the same criteria. In addition to surfacing more helpful commentary from users, the update should help shoppers learn about changes made by manufacturers to existing products that aren’t officially announced or mentioned in their listing. There’s no word yet on whether Amazon will roll this update out to other countries. We’ve contacted the company to find out and will update this post when we hear back.