AI Insights
Understand how the recommendation engine works
Scoring Strategies
Genre First
ActivePrioritizes matching your preferred genre with high weight on genre similarity.
Mood First
Prioritizes mood alignment. Best when you need music for a specific feeling.
Energy Focused
Prioritizes energy level match. Ideal for workouts, studying, or relaxing.
Weight Distribution
Confidence Distribution
Distribution of confidence scores across your recommendation history.
AI Knowledge Base
The retrieval layer uses these 6 knowledge notes to provide supporting evidence for recommendations and boost confidence scores.
Genre familiarity
Pop and indie-pop songs usually perform well for users who ask for upbeat and high-valence tracks.
Focus listening
Lofi tracks with moderate energy and higher acousticness are often preferred for focus and study contexts.
High intensity
High-energy rock and metal songs are reliable matches when users request intense listening sessions.
Acoustic preference
Users who explicitly prefer acoustic sounds usually respond better to songs with high acousticness and softer mood tags.
Dance/party profile
Party-oriented profiles correlate with high danceability and euphoric mood tags.
Diversity reminder
Recommendation quality improves when repeated artists and genres are penalized to avoid narrow result sets.