Can Machine Learning Improve Online Reviews?


Having assisted in everything from search engines and sentiment analysis to fraud detection and medical diagnosis, machine learning is getting a new use: diminishing the power of fake online reviews.

Amazon plans on deploying a machine-learning platform to weight certain reviews on its site as more important than others, according to CNET. Reviews written by verified purchasers, for example, will have more influence, along with those judged “helpful” by the overall community. “It improves over time,” an Amazon spokesperson told the publication.

Amazon has toyed quite a bit with machine learning over the past year. In April, for example, it launched a cloud-based machine-learning service, complete with APIs and Machine Learning (ML) models. That service enables developers with relatively little background in machine learning to build complex algorithms without needing to worry about coding or infrastructure.

Check out the latest machine learning jobs.

As machine learning becomes more prevalent (and cheaper), it’s likely that more companies will use it to improve even relatively mundane and straightforward products and features, such as product reviews. For those who specialize in machine learning, that can open up new career opportunities.