How Pinterest Uses Machine Learning To Keep Its Users Pinned

Five ways the image-sharing site is harnessing AI to keep people engaged.

How Pinterest Uses Machine Learning To Keep Its Users Pinned
[Illustration: Amrita Marino]

Thanks to recent gains in machine learning, computers are getting skilled at picking out patterns and features in text and images. That’s how e-commerce giants like Amazon and eBay build sophisticated recommendation systems and how social networks like Facebook and Twitter are tweaking feeds to keep users hooked. Pinterest is no exception, with 30% of engagement tied to personalized real-time suggestions. Here’s how Pinterest engineers are leveraging artificial intelligence to keep the website’s 150 million–plus users pinning and sharing.


Identifying Visual Similarities

Machine learning can not only determine the subject of an image, it can also identify visual patterns and match them to other photos. Pinterest is using this technology to process 150 million image searches per month, helping users find content that looks like pictures they’ve already pinned. Pin a photo of a cheetah-print pillow, and Pinterest will serve up animal-print decor from other users. Future iterations of the Pinterest app may let users simply point their cameras at real-world objects to get instant recommendations.

Categorizing And Curating

If a user pins a mid-century dining-room table, the platform can now offer suggestions of other objects from the same era. The key? Metadata, such as the names of pinboards and websites where images have been posted, helps the platform understand what photos represent.

Predicting Engagement

While many platforms prioritize content from a user’s friends and contacts, Pinterest pays more attention to an individual’s tastes and habits—what they’ve pinned and when—enabling the site to surface more personalized recommendations. After all, friends who like the same recipes may not agree at all on fashion.

Prioritizing Local Taste

Pinterest is an increasingly global platform, with more than half of its users based outside the U.S. Its recommendation engine has learned to suggest popular content from users’ local region in their native language. One finding: Slow-cooker recipes are more popular in the U.S. than the U.K., where the appliances aren’t as common.

Going Beyond Images

Analyzing what’s in a photo is a big factor in the site’s recommendations, but it doesn’t offer the whole story. Pinterest also looks at captions from previously pinned content and which items get pinned to the same virtual boards. That allows Pinterest to, say, link a particular dress to the pair of shoes frequently pinned alongside it, even if they look nothing alike.

About the author

Steven Melendez is an independent journalist living in New Orleans.