Filter Bubble / Echo Chamber

Algorithmic Recommendation Systems

Categories:

What you have seen, listened to, bookmarked, liked, or evaluated positively

Audio Features – measurable acoustic properties of tracks:

• Danceability, Energy, Loudness, Tempo, Valence, Acousticness, Instrumentalness, Speechiness, Liveness, Duration, Key, Mode.
• Genre, Artist, Album, Release year, Language, Popularity index.

User Behavior – interaction patterns:

• Play count, Skip rate, Playlist additions, Listening duration, Repeat rate, Like/Dislike actions, Session time, Device type.

Contextual Metadata – listening context:

• Time of day, Location, Device, Mood tags, Current activity (e.g. “workout,” “focus”).

Collaborative Signals – similarity inferred from collective usage:

• Users with similar tastes, Shared playlist inclusion, Co-listening patterns.

Editorial & Social Categories – human-curated or social data:

• Official playlists, Influencer playlists, Follower relations, Artist follows.

Temporal Trends – recency and evolution of taste:

• Short-term, medium-term, and long-term preference windows.

What others like you have liked

POLARIZATION (Scott Galloway)