In the previous article, I explained how TikTok recommends videos in the For You tab and, in this, the idea is to learn some tricks to hack the algorithm. The initial feed acts as a discovery area and is also where most users spend their time. Seeing what you “don’t like” in the feed is a sign that TikTok didn’t understand your tastes. However, it is possible to give a little help to the algorithm and tame the suggestions on the home screen.
New accounts may spoil
New users, who created accounts more recently, are invited to select interest categories, such as pets or travel, to help adapt the For You tab recommendations to their preferences. This allows the app to develop an initial feed as soon as you start using it, and then the algorithm will polish the recommendations based on interactions with that initial set of videos.
For old users or those who have not selected categories, TikTok does the service and offers a generic feed of popular videos to start suggesting content.
Your first set of likes, comments and replays will start an initial round of recommendations as the system learns more about your taste. So if you watch, like, comment and send this video to others, it can help.
How to find more of what I like?
Here’s the mystery of TikTok …
Each new interaction helps the system to know your interests and suggest content – so it seems obvious, the best way to improve or adapt your feed is to simply use and enjoy the app interacting with what you like. It’s easy to get lost in the initial tab and review videos that you “didn’t like” or “didn’t understand”…
Over time, “For You” should be able to make recommendations more relevant to your interests. But any slippage makes the algorithm value other videos.
It is worth noting that the For You feed is not only shaped by interaction through the feed itself. When you decide to follow new accounts, for example, this action also helps to refine your recommendations (from similar new accounts), in addition to exploring hashtags, songs, effects and trending topics on the “Discover” tab. All of these are ways to personalize your experience and attract new categories to your feed.
How to stop seeing what I don’t like?
Literally, stop seeing what you don’t like!
I’ll tell you a secret: if you press a video for a few seconds, you can add it to your favorites or just tap “Not interested”.
You can also hide videos from a specific creator or made with a certain sound (music) or report a video that appears to be at odds with the guidelines. All of these actions contribute to future recommendations in your feed.
Some like bubbles…
One of the challenges with any recommendation engine is that they can limit experience – what is sometimes called a “bubble”. According to TikTok, to keep For You interesting and varied, the system works to merge several types of new content together with those you already know and like.
For You, for example, tends not to show two videos in a row made with the same song or by the same creator. It also does not recommend content that you have seen before or any content that is considered spam. However, a video that has become popular with other users with similar interests can be recommended.
“Diversity is essential and you can sometimes find a video in your feed that doesn’t seem relevant to your interests or that hasn’t accumulated a large number of likes. This is an important and intentional component of our approach to recommendation: bringing a diversity of videos to For You offers additional opportunities for you to come across new categories of content, discover creators and experience perspectives and ideas as you browse your feed”, The social network explains.
By offering different videos from time to time, the system can also get a better sense of what is popular and what is not. That is, something can only be popular with fans of a specific creator and not a viral one, in fact, that pleases all types of users.
Finally, TikTok claims that the recommendation system was also designed with security in mind. The content is reviewed and some types of videos may not be eligible. Likewise, videos that have just been uploaded or are still being reviewed (or, worse, are spam) also do not enter this initial tab.
But everything changes, all the time
None of this, however, is fixed.
Everything changes all the time. The development and maintenance of the TikTok recommendation system – that is, the algorithm – is an ongoing process.
“We work to refine accuracy, adjust models and reassess the factors and weights that contribute to recommendations based on user feedback, research and data. We are committed to more research and investment, while working to avoid bias that can affect any recommendation system ”.
With information: TikTok