About Revibe
Revibe is an intelligent music recommendation system that helps you discover new music through advanced audio analysis and machine learning technology.
Machine Learning Powered
Using K-means clustering with distinct clusters, we group songs based on their audio features to find the most similar tracks.
Audio Analysis
We analyze multiple audio features including danceability, energy, valence, acousticness, and tempo to understand each song's unique characteristics.
Vast Music Database
Our system is built on a comprehensive database of songs, allowing for diverse and accurate recommendations across different genres and styles.
Personalized Discovery
Find new music based on songs you already love, helping you discover artists and tracks that match your musical taste.
How It Works
Revibe uses a sophisticated K-means clustering algorithm with 20 clusters to group songs based on their audio features. Our recommendation process follows these steps:
- Standardizes the song's audio features using StandardScaler
- Identifies which of the 20 clusters the song belongs to
- Finds songs within the same cluster
- Calculates Euclidean distances to find the most similar matches
- Enriches results with iTunes preview data
The audio features we analyze include:
- • Acousticness
- • Danceability
- • Energy
- • Instrumentalness
- • Liveness
- • Loudness
- • Speechiness
- • Tempo
- • Valence