About YouTube recommendation system

YouTube shared some very interesting and important insights about their recommendation system,so that's what we're going to talk about in a summarized way.

We always talk about how important a few factors are to show YouTube that an EXO video isworth being recommended to more viewers. So let's understand a little bit more about what the recommendation system is and how it works.

WHAT IS A RECOMMENDATION SYSTEM?

The YouTube recommendation system is designed to help users find videos that they want to watch and that will provide them with value. There are two main places where recommendations can be found: on the user's homepage and in the "Up Next" panel while watching a video.

In the early days of YouTube, the recommendation system ranked videos based on popularity, but now it uses a user's unique viewing habits to suggest other content they may want to watch. YouTube's system compares a user's viewing habits with those who have similar habits and recommends content based on this comparison. The success of YouTube's recommendations depends on accurately predicting the videos users want to watch, rather than connecting them through social networks. Users have control over how much data they share with YouTube and can pause, edit, or delete their search and watch history whenever they want.

HOW YOUTUBE PERSONALIZE RECOMMENDATIONS

YouTube's recommendation system is personalized and constantly evolving by learning from over 80 billion pieces of information, including clicks, watchtime, survey responses, sharing, likes, and dislikes.

• Clicking on a video provides an indication of what you might want to watch, but watchtime gives a more personalized signal.

• Survey responses help to measure "valued watchtime," which is the time spent watching a video that you consider valuable.

• Sharing and liking videos also provide signals that can predict the likelihood of sharing or liking further videos.

All of these signals are important, but their importance varies depending on the individual. YouTube's recommendation system develops dynamically based on viewers' habits.

For the full article, read on YouTube's official blog: https://blog.youtube/inside-youtube/on-youtubes-recommendation-system/.