Your taste tribe exists

Beyond simple "similar users"

Most platforms show you users who liked the same popular things. That's not very useful — everyone likes The Shawshank Redemption. CatNLog's communities are different. They're based on your entire taste profile, including the unpopular opinions that make your taste unique.

The user who shares your love of both obscure 70s prog rock AND cozy mystery novels AND survival horror games? That's a real taste-twin. Their recommendations will be far more valuable than someone who just happens to like the same Top 10 films.

How clustering works

K-means clustering is an algorithm that groups data points — in our case, user taste profiles — into clusters based on similarity. Each user's profile is a vector of preferences across hundreds of dimensions (genres, themes, styles, eras, and more).

The algorithm iteratively assigns users to the cluster whose center (centroid) is closest to their profile, then recalculates the centers. After convergence, users in the same cluster share the most similar taste profiles. Your community is your cluster.

Living communities

As you add more ratings, your taste profile evolves — and so does your community membership. Rate a bunch of jazz albums and you might drift toward a community that values musical complexity. Start exploring Korean cinema and find yourself among world cinema enthusiasts. Your community always reflects your current taste.