The Algorithm Knows You... Or Does It?

YouTube feels infinite — millions of creators, endless content, and a recommendation system deciding what you see next. But behind the algorithm, something much more human is happening.

Every comment is a trace of movement, a footprint. A tiny sign that someone chose to stop, react, connect, or speak.

So what happens if we stop looking at YouTube as a platform of videos… and start looking at it as a universe of people?

We will follow these footprints to uncover how channels connect, and what kinds of people bring them to life.

Two Perspectives. One Question.

Using the Youniverse dataset — more than 8.6 billion comments written by tens of millions of users — we explore a simple but profound question:

What does YouTube really look like when we map it through human behavior?

This question feels like a lot. We decided to split it into multiple research questions.

Our research questions

User Behavior: Is commenting structured enough to build a meaningful network?
Organic Structure: What structures emerge when we connect channels through people, not algorithms?
Recommendation: Can we build a transparent recommendation engine based on human interaction?

The path to answer these questions led us to two complementary stories.

Two Ways to Explore the Youniverse

The Network Journey
In this path, we guide you. We begin by carefully cleaning and validating billions of comments to isolate genuine human engagement. From there, we connect channels through shared audiences, revealing the large-scale structure of YouTube — its hubs, bridges, and hidden communities — as seen through an analyst’s lens.
The User Journey
In this path, you take over. Instead of starting from predefined labels, you step back and decide how behavior itself should be compared. By grouping users with identical commenting patterns, you explore recurring ways people actually interact with YouTube — discovering structure not because we point it out, but because you observe it emerge.

Why Both Stories Matter

These two perspectives reinforce each other.

The network view shows the large-scale architecture of YouTube built from shared engagement.
The user view shows the individual behaviors that create that architecture in the first place.

Together, they reveal that YouTube is not organized by algorithms alone; it is shaped by people, their routines, their curiosity, and their conversations.