Recommended Videos for You
- Recommendation systems serve as the primary mechanism for content discovery across modern digital platforms, including streaming services, search engines and social media.
- The functionality of these systems is evident in curated feeds, such as those found on Facebook Reels or the recommendation lists that suggest videos based on duration and...
- TikTok utilizes a recommendation system to power its For You feed, which is designed to deliver content likely to be of interest to a specific individual.
Recommendation systems serve as the primary mechanism for content discovery across modern digital platforms, including streaming services, search engines and social media. These systems are designed to provide a personalized experience by suggesting content based on user preferences and interactions.
The functionality of these systems is evident in curated feeds, such as those found on Facebook Reels or the recommendation lists that suggest videos based on duration and view counts to keep users engaged.
TikTok Recommendation Mechanics
TikTok utilizes a recommendation system to power its For You
feed, which is designed to deliver content likely to be of interest to a specific individual. The platform maintains that there is no single For You
feed; instead, each user’s experience is unique and tailored to them.
The system ranks videos using a combination of factors. For new users, the system starts with the interests they express. As the user interacts with the app, the system adjusts recommendations based on signals that indicate whether a user is or is not interested in specific content.
Specific user interactions that help the system gauge preference include:
- Following an account
- Posting a comment
- Indicating a lack of interest in specific content
YouTube User Controls and Tuning
YouTube provides users with various tools to influence their recommendations and search results. Users can manage their experience by removing specific videos from their watch history or deleting searches from their search history.
One primary method of tuning recommendations is providing direct feedback on the Home and Watch Next pages. Users can mark a video as Not interested
to help the system improve future suggestions.
When marking a video as Not interested
, YouTube allows users to provide further detail to customize the recommendation engine. Available options include:
- Indicating that the user has already watched the video
- Stating that the user does not like the video
- Selecting
Don’t recommend channel
users have the option to turn off their watch and search history entirely. Deleting and disabling the watch history removes video recommendations from the Home feed.
For users on Smart TVs, streaming devices, or game consoles, these controls are accessible via the remote, where pressing and holding the select button on a recommended video allows them to mark it as Not interested
and share the reason why.
