TikTok revealed some details about how its algorithm works

Recently The news was released that TikTok revealed some of its internal mechanisms of its algorithm "Highly valued" that keeps hundreds of millions of users around the world hooked on the video application.

And it is that the algorithm is classified as highly appreciated because it is this code that has allowed it, in just 2 years of existence, to get close to 700 million users and become a company with a remarkable valuation between 20 to $ 30 thousand millions.

This is all because ByteDance, the parent company of TikTok, faces one of the toughest court mandates in recent years.

Well basically the president of the United States, Donald Trump, is forcing him to sell US shares of TikTok to a local company.

The deadline is set for September 15 and we still only hear rumors of a possible sale to Microsoft.

Additionally, Trump said Thursday that he did not plan to extend the deadline for ByteDance to sell the US branch of TikTok.

However, while an important ByteDance asset is at stake, TikTok, for its part, tries to shed light on the code of its algorithm. During a call with Axios reporters on Wednesday, TikTok executives said they were revealing details of their algorithm and data practices to dispel some myths and rumors about the company.

In fact, TikTok executives offered journalists a virtual tour for his new "transparency center" in Los Angeles, before talking about the algorithm itself.

When asked how the algorithm works, TikTok executives said the algorithm uses machine learning to determine which content an app user is most likely to interact with and serve them over and over again, looking for videos that are similar or that people with similar user preferences like.

In more detail, they said that When users open TikTok for the first time, they watch 8 popular videos showing different trends.

After this first contact, the algorithm will continue to serve to the user with more iterations of 8 videos making reference to the first videos with which the user has interacted and what they are doing.

It also identifies videos similar to those that attracted the user through the video information, which can include details such as subtitles, hashtags or sounds. The recommendations also take into account device and user account settings, which include information such as preferred language, country, and device type.

Once TikTok has collected enough user data, the app can map a user's preferences to similar users and group them into "groups."

At the same time, he also groups the videos into "groups" based on similar themes, such as basketball or rabbits. Additionally, through machine learning, the algorithm delivers the videos to users based on their proximity to other user groups and the content they like.

TikTok logic aims to avoid redundancies that could annoy the user, such as watching multiple videos with the same music or from the same creator.

According to TikTok, its ability to identify user preferences is due to the fact that its algorithm can produce "filter bubbles", reinforcing the preferences of existing users rather than simply showing them more varied content, expanding their content, horizons or offering them opposing points of view.

These include their lifespan and how a user finds them, to further break them down if necessary. Company executives added that because filter bubbles can be used to reinforce conspiracy theories, hoaxes and other misinformation, TikTok's product and policy teams study which accounts and video information (topics, hashtags, captions) , sound, etc.) are probably related to misinformation. Then moderation measures are taken.

In addition, the briefing also provided an update on TikTok's security, privacy and data collection practices. The company says it tries to resolve and prevent incidents on its platform before they happen.

The content of the article adheres to our principles of editorial ethics. To report an error click here!.

A comment, leave yours

Leave a Comment

Your email address will not be published.



  1. Responsible for the data: Miguel Ángel Gatón
  2. Purpose of the data: Control SPAM, comment management.
  3. Legitimation: Your consent
  4. Communication of the data: The data will not be communicated to third parties except by legal obligation.
  5. Data storage: Database hosted by Occentus Networks (EU)
  6. Rights: At any time you can limit, recover and delete your information.

  1.   Oscar Reyes said

    The wallpaper they used to tick toc looks like the official Ubuntu background