Chapter
Improving YouTube Recommendations with Experiments
1:05:56 - 1:10:54 (04:58)
The success of YouTube's recommendation experiment can be determined by the decrease in video dismissals and an increase in viewer satisfaction and five-star ratings. The first diversity measure was limiting the number of videos from the same channel in a row, but viral videos now play a part in improving recommendations.
Clips
YouTube experiments run for weeks to months and measure hundreds of variables to ensure that changes made to the algorithm are successful and improve the viewing experience for users.
1:05:56 - 1:08:43 (02:47)
Summary
YouTube experiments run for weeks to months and measure hundreds of variables to ensure that changes made to the algorithm are successful and improve the viewing experience for users. Refining the system over time allows it to better learn how to handle different situations based on past observations.
ChapterImproving YouTube Recommendations with Experiments
EpisodeCristos Goodrow: YouTube Algorithm
PodcastLex Fridman Podcast
In this episode, the speaker discusses the phenomenon of viral videos on YouTube and how the platform recommends popular videos to its users based on their interests.
1:08:43 - 1:10:54 (02:11)
Summary
In this episode, the speaker discusses the phenomenon of viral videos on YouTube and how the platform recommends popular videos to its users based on their interests. They also touch on the topic of predicting when a video will go viral.