TikTok's recommendation system, "For You," was recently explained by the social media platform. According to Techcrunch's latest report, when a user opens the short-form video app, they are met with a stream of popular videos tailored to their interests.
However, the content you're seeing may be different from the ones that your friend or other people see when they use the app. Since the videos provided are based on how the users personalized their app, TikTok named the mainstream content "For You."
Also Read: [CHINA Vs. INDIA] TikTok, Xiaomi and other Chinese Apps Flagged by Indian Intel Agencies: Here's Why
However, it is still unknown how the recommendation system works on the social media platform. Tiktok explained on its own website how its algorithm works, detailing the factors involved, as well as how they are weighted for each individual user.
According to TikTok's report, filter bubbles or feeds where users are presented with a homogeneous stream of videos are not created by the system.
Also Read: EA's New Rules Included in the Positive Player Charter Initiative Helps Combat Racism, Homophobia, and Sexism
Just like many recommendation systems, the "For You" feed is also powered by user input. In TikTok's case, to help the users determine their interests, the system takes into account the videos they've liked or shared, the accounts they follow, the comments they post, and also the content they create.
How is TikTok's algorithm different from Instagram, Facebook, and YouTube
Just like users' language preference, device type, and country setting, TikTok also uses the device and account settings to tailor content for each individual user. However, compared to other recommendation systems, TikTok said that these factors receive a lower weight in the recommendation system since they're focused on making sure the system's performance is optimized.
TikTok explained that the social media platform's understanding of what a user likes is also refined by other signals such as the time spent by a user watching a video which indicates their interests. The recommendation system will also affect the video's information such as sounds, hashtags, and captions associated with the user's preferred content.
The social media platform will also provide the user's "For You" feed with content that may be different from their interests. This feature is used to add diversity by allowing users to encounter new creators and new content categories, giving them new ideas and perspectives.
It could be argued that other platforms such as Instagram, Facebook, and YouTube haven't totally addressed the issue in content recommendation diversity. The algorithms of these big social media platforms highlight more of the same sort of content the users previously liked, keeping them in their own echo chambers.
