Revealed: The reason you see customized content on your Instagram Explore Tab

Have you ever noticed that the Explore tab on your Instagram app has posts that are very similar to your interests? Interestingly, no two Explore tabs are the same and the posts that you see on your mobile are poles apart from your friend.

This is because Instagram integrates an advanced artificial intelligence (AI) system that is built on a 3-part ranking funnel. The AI system has the capability to extract 65 billion features and makes 90 million model predictions every second.

Of course, it was apparent that Artificial Intelligence was behind the personalized Explore Tab. However, Facebook, Instagram’s parent company explained the technical features of the AI system in a detailed blog post published on Monday.

Facebook also said that the experience with the Explore tab could be enhanced by interaction. For example, if a user sees posts that they don t like – they can click on the three dots in the top right corner and click on ‘see fewer posts like this.’

On the other hand, simply liking, commenting, sharing, and expanding the posts in the Explore tab will indicate the algorithm to send more of such content your way.



According to the Facebook engineers, the AI uses a method called ‘word embedding.’ This method studies the order in which words appear with other words to determine the accounts similar to each other. For example, words like ‘birthday’, ‘cakes’, and ‘party’ are usually co-related. However, they will not appear together with ‘cat’ and ‘library’.


Moreover, the algorithm also records the user’s interaction with different posts and provides them with the best possible matches. In fact, Facebook explains that the AI narrows down from the billions of posts and comes up with the top 25 posts for your Explore grid.

Facebook further added that the AI filters for inappropriate content as well. The leading social media company said that they want to provide content that is both –safe and appropriate for the community. This is why they use a variety of methods to filter out the ‘unsafe’ content from their database. Additionally, the company constantly works on innovative methods to improve their AI system.
"The scale of both the Instagram community and inventory requires enabling a culture of high-velocity experimentation and developer efficiency to reliably recommend the best of Instagram for each person’s individual interests. Our custom tools and systems have given us a strong foundation for the continuous learning and iteration that are essential to building and scaling Instagram Explore.", explained Facebook.
Read next: Instagram’s Spark AR Now Has Tracking Effect And Slider Controls To Give Tough Time To Snapchat
Previous Post Next Post