How TikTok reads your mind – Community News
Us China

How TikTok reads your mind

“This system means viewing time is essential. The algorithm tries to make people addicted instead of giving them what they really want,” said Guillaume Chaslot, the founder of Algo Transparency, a group in Paris that has studied YouTube’s recommendation system and has a dark view of the effect of the algorithm. product on children in particular. Mr Chaslot has reviewed the TikTok document at my request.

“I think it’s a crazy idea to let TikTok’s algorithm control our children’s lives,” he said. “Every video a child watches gets a piece of information about him on TikTok. Within a few hours, the algorithm can detect his music taste, his physical attraction, whether he is depressed, whether he takes drugs, and many other sensitive information. There is a high risk that some of this information will be used against him. It could potentially be used to micro-target him or make him more addicted to the platform.”

According to the document, watch time is not the only factor TikTok considers. The document provides a rough comparison for how videos are scored, summarizing a machine learning prediction and actual user behavior for each of the three bits of data: likes, comments, and playtime, as well as an indication that the video played:

Plike X Vlike + Pcomment X Vcomment + Eplaytime X Vplaytime + Pplay X Vplay

“The recommendation system scores all videos based on this comparison and returns videos from users with the highest scores,” the document says. “For brevity, the comparison has been greatly simplified in this document. The actual equation used is much more complicated, but the logic behind it is the same.”

The document illustrates in detail how the company is tweaking its system to identify and suppress “like bait” — videos designed to play up the algorithm by explicitly asking people to like them — and how the company is considering more nuanced questions. .

“Some authors may have cultural references in their videos, and users can only better understand those references by watching more of the author’s videos. Therefore, the total value a user watches all those videos is greater than the values ​​of watching each individual video added together,” the document said. “Another example, if a user likes a certain kind of video, but the app keeps pushing the same kind to them, they would quickly get bored and close the app. In this case, the total value created by the user watching the same type of videos is lower than that of watching each individual video, because repetition leads to boredom.”

“There are two solutions to this problem,” the document continues. “Make some assumptions and divide the value into the value equation. For example, in terms of repeated exposure we can add a value ‘same_author_seen’ and for the problem of boredom we can also add a negative value ‘same_tag_today’. Other solutions besides value comparison can also work like forced recommendation in users for u feed and dispersion etc. For example, the boredom problem can be solved by dispersion.

About the author


Add Comment

Click here to post a comment