Buh bye fakers? Amazon tweaks customer product reviews system

Retail giant hopes machine learning will weed out tricksters

Snake oil salesman

Amazon has rejigged its product reviews systems for customers in the US, in an apparent move to mute fake ratings on the retail giant's service.

The company said it was now using machine-learning tech that it had developed to try to flush out old and crappy reviews, according to Cnet.

Amazon's Julie Law was quoted as saying:

The system will learn what reviews are most helpful to customers ... and it improves over time. It's all meant to make customer reviews more useful.

Presently, the software will only be deployed Stateside, the report said. Amazon's overhauled system will rank newer reviews higher and give more weight to its verified buyers.

The multinational will push up popular reviews that have been flagged up as helpful by customers. Star ratings will also get a boost under the new system, apparently.

Amazon has long been blighted by bogus reviews. But it's not a huge fan of acknowledging the problem, as this Reg story – published way back in 2000 – illustrates. ®




Biting the hand that feeds IT © 1998–2018