The Chinese social network TikTok revealed some of the secrets on how its video recommendation algorithm works in the ‘For you’ section of the application, where it collects the elements it takes into account so that users can access new content.
By opening the ‘For You’ section of the short video app, users can watch a series of selected short films based on their interests, making it easier to search for content.
The system recommends content based on a series of factors, starting with the interactions of users, such as the content they like or share, the accounts they follow, among others.
TikTok noted that the recommendation system is also based on information from videos, such as subtitles, sounds, and ‘hashtags’, as well as device and account settings, such as language preference, country settings, and type of device; the latter factor to a lesser extent.
Likewise, the system records as a stronger indicator of interest the fact that a user watches a video from start to finish instead of going to the next. The shorts are then ranked to determine the probability that a user is interested in content.
Users can also prevent the application from recommending a series of videos by long pressing on one and selecting that they are not interested in that particular one. They can also hide videos from a specific creator or with a specific sound.