Meta explains how artificial intelligence ranks content on Facebook and Instagram
Meta has revealed how its artificial intelligence determines which posts to show on Facebook and Instagram.
Here's What We Know
The tech giant has released 22 cards explaining how its systems rank content for Ribbon, Stories and Reels. Each card describes how a particular aspect of the platform works.
For example, for the Facebook Ribbon, the system collects all possible posts from friends and Pages to which the user is subscribed. The AI then takes into account different cues, such as the author of the post, interactions with it and the number of likes from friends. These aspects help the system rank the posts in terms of relevance and value and show them in the feed in descending order of scores.
The cards also show how content can be customised. In addition, the company has developed instructions on how to unsubscribe from a person or group or use the "show more" or "show less" buttons on a post to indicate relevance to a particular topic.
Meta's President of Global Affairs Nick Clegg listed tools to personalise the experience on the company's platforms. He recalled that Facebook and Instagram have centralised menus with customisation settings.
Clegg also said that Meta is testing a new feature for Reels that will help improve video recommendations. The company is also working on making "show more" and "show less" buttons more prominent.
Source: Engadget.