How YouTube recommendations or data personalization work
Feb 24, 2024 23:10:04 GMT -5
Post by account_disabled on Feb 24, 2024 23:10:04 GMT -5
Version: Olga Kotsofane. Author: Oksana Kharkovchuk YouTube recommendations connect billions of users around the world with meaningful, educational and entertaining content. At the same time, it is very important that the recommendations are properly configured, since their task is to find the right audience for each video (referring viewers and bringing high views). About thistoldon Google's blog, Christos Goodrow, YouTube's vice president of technology. You can find recommendations on the main page of YouTube (personalized recommendations, subscriptions and news) and in the " Next " panel (offering additional content while watching videos).
The system compares the user's actions with similar behavior and, as a result, offers content that will interest you. To personalize YouTube Business Owner Phone Numbers List recommendations, the system uses the rules and "signals" of 80 billion information objects. These can be clicks, viewing time, links to content, survey results, likes and dislikes — signals that can determine your preferences. Let's consider them in more detail: Read also : How to Create YouTube Ads: 5 Tips to Improve Your Strategy by Ben Jones To personalize YouTube recommendations, the system uses the rules and "signals" of 80 billion information objects. These can be clicks, viewing time, links to content, survey results, likes and dislikes — signals that can determine your preferences.
Let's consider them in more detail: Clickability means that, most likely, the video content seems interesting to you, taking into account the time of the watched clips; Taking into account viewing time allows you to get personalized signals about what content you are most likely to enjoy. The benefit of watching different videos can vary, so it is important for YouTube to analyze “ valuable watch time ” and viewer satisfaction; " Valuable viewing time " is evaluated thanks to the results of surveys (video stars on a scale of one to five), which allows you to determine how satisfied viewers are with the content. In personal recommendations, only those videos that you gave 4 or 5 stars are taken into account. It is also possible to predict ratings from YouTube excluding those already posted by users; When you “ Like ” or “ Dislike ” a video , share a video or link to content, you are giving your actions a certain characteristic.
The system compares the user's actions with similar behavior and, as a result, offers content that will interest you. To personalize YouTube Business Owner Phone Numbers List recommendations, the system uses the rules and "signals" of 80 billion information objects. These can be clicks, viewing time, links to content, survey results, likes and dislikes — signals that can determine your preferences. Let's consider them in more detail: Read also : How to Create YouTube Ads: 5 Tips to Improve Your Strategy by Ben Jones To personalize YouTube recommendations, the system uses the rules and "signals" of 80 billion information objects. These can be clicks, viewing time, links to content, survey results, likes and dislikes — signals that can determine your preferences.
Let's consider them in more detail: Clickability means that, most likely, the video content seems interesting to you, taking into account the time of the watched clips; Taking into account viewing time allows you to get personalized signals about what content you are most likely to enjoy. The benefit of watching different videos can vary, so it is important for YouTube to analyze “ valuable watch time ” and viewer satisfaction; " Valuable viewing time " is evaluated thanks to the results of surveys (video stars on a scale of one to five), which allows you to determine how satisfied viewers are with the content. In personal recommendations, only those videos that you gave 4 or 5 stars are taken into account. It is also possible to predict ratings from YouTube excluding those already posted by users; When you “ Like ” or “ Dislike ” a video , share a video or link to content, you are giving your actions a certain characteristic.