Recently, everyone seems to be complaining about algorithms. How exactly do recommendation systems work? For example, is watching a single movie enough, or does just clicking the like button suffice? Or do they base it on user behavior? Do they perceive user preferences as flexible or rigid? What have you all noticed?
How do digital platforms recommend content?
👁️ 18 views💬 1 replies❤️ 0 likes
1 Replies
Bro, the algorithms on digital platforms are so complicated that you can’t even ask “is just watching one movie enough” anymore. I recently watched just one series on Netflix, and then for three months it kept recommending that same kind of content, I swear. They’re not just looking at what you watch; they even track how long you skip and where you quit. For example, if you left the last episode halfway, they code the perception that “this person can’t finish the show.”
But the weirdest part is that hitting the like button locks the system in the same way. I think the impact of that button is so strong that, say, you enjoyed a movie for only 30 seconds and they keep pumping out that type of content. I once made a mistake and hit “like” on a movie I hated, and after that, that kind of stuff kept showing up. I think we realize how flexible the algorithms are the most when users change. For instance, I wanted to rewatch a K‑drama I saw in November in February, but the system cut off recommendations in that category, saying “you won’t be watching a series tonight.” So in the end they can stretch based on our behavior, but that initial rigidity stays the same.