posts / Humanities

The Librarian Who Reads Your Preferences

phoue

4 min read --

Ji-hye’s Happy Dilemma

There’s a newly opened library in a small town. Ji-hye, the librarian there, has one goal: to help every visitor discover their own “life-changing book” before they leave.

But everyone’s tastes are so different. What book should she recommend to truly capture their hearts? Ji-hye pondered while flipping through the library’s loan record cards. And there, she discovered two fascinating patterns. These patterns are the core principles of what we call Collaborative Filtering today.


First Discovery: Finding Your “Soulmate” with Similar Tastes

The Connection Between Min-jun and So-ra

Ji-hye was surprised when she noticed that Min-jun and So-ra had borrowed very similar lists of books. Both loved fantasy novels and enjoyed historical mysteries.

One day, Min-jun borrowed a newly arrived fantasy novel and said, “This book is amazing!” Hearing this, Ji-hye immediately thought of So-ra.

“Ah! If Min-jun liked it so much, So-ra will surely love it too!”

The next time So-ra visited the library, Ji-hye recommended the book without hesitation, and So-ra became an enthusiastic fan as well.

This is ‘User-Based Collaborative Filtering’

What Ji-hye just did is the basic principle of User-based Collaborative Filtering. Simply put, it recommends “what other people with tastes similar to yours liked.”

  • Netflix Example: Suppose I enjoyed the thriller movie and the sci-fi movie . Netflix finds many others who also rated both highly. If those people also liked another movie, say , Netflix will recommend it to me. Even if I’ve never heard of it, since people with similar tastes have already vetted it, the chance I’ll like it is high.
  • YouTube Example: Suppose I often watch “cat” videos and subscribe to “game streaming” channels. YouTube finds other users who enjoy both “cats” and “games.” If that group recently started watching a lot of “cooking” videos, then suddenly a “Baek Jong-won recipe” video might appear on my YouTube homepage. This shows the power of user-based filtering, which recommends new interests based not only on my behavior but also on the behavior of users with similar tastes.

Second Discovery: Finding “Best Friend Books” That Attract Each Other

The Secret of and

Ji-hye noticed that people who borrowed often borrowed shortly after. Conversely, those who were moved by often sought out .

It was as if the two books were saying, “We’re perfect friends!”

So when someone borrowed , Ji-hye naturally asked, “Have you read Paulo Coelho’s ? If you liked this one, you’ll definitely enjoy that too.” Surprisingly, most of these recommendations were successful.

Advertisement

This is ‘Item-Based Collaborative Filtering’

Ji-hye’s second discovery illustrates the principle of Item-based Collaborative Filtering. This time, the focus is not on people but on the relationships between items—content. It recommends “other things similar to what you liked.”

  • Netflix Example: Netflix’s “More Like This” feature is a prime example. If I binge-watch , Netflix analyzes what other shows people who watched also watched. It might find that or are frequently watched next, so it recommends those. This isn’t because of shared directors or actors, but purely based on consumption pattern data revealing strong connections between the content.
  • YouTube Example: YouTube uses item-based filtering in the “Up Next” list on the right or recommended videos below. For example, if I watch “IU’s live clip,” the system analyzes what other videos viewers of that clip watched next—like “Taeyeon’s live clip” or “covers of IU songs by other singers”—and recommends those. The “IU live” item and the “Taeyeon live” item have become best friends in users’ viewing histories.

The Invisible Librarian Among Us

Smart librarians like Ji-hye are everywhere around us. Netflix and YouTube use a sophisticated mix of these two collaborative filtering methods. Sometimes they find your “soulmate” with similar tastes and borrow their choices (user-based), and other times they introduce “best friend” content related to what you’ve watched (item-based).

So next time you hear the term “collaborative filtering,” why not think of the friendly librarian Ji-hye who finds someone with tastes like yours or a best friend book for the one you loved? Though the term may sound technical and cold, its essence began with a warm heart aiming to connect people with joy and satisfaction.

#Collaborative Filtering#Recommendation Systems#Netflix Algorithm#YouTube Algorithm#User-Based Filtering#Item-Based Filtering#Artificial Intelligence

Recommended for You

Autonomy Premium: How to Buy Back Your Time with Money, You Too Can Become Truly Wealthy

Autonomy Premium: How to Buy Back Your Time with Money, You Too Can Become Truly Wealthy

14 min read --
How Amazon and Google Designed Failure to Achieve Success

How Amazon and Google Designed Failure to Achieve Success

11 min read --
Why Does a Rising Salary Not Bring Happiness? The Secret to Becoming 'Rich in Time'

Why Does a Rising Salary Not Bring Happiness? The Secret to Becoming 'Rich in Time'

7 min read --

Advertisement

Comments