Music Mood Analysis

The Music Mood Analysis and Spotify Wrapped Predictor is a Python-based project designed to analyze users' listening habits and predict their music preferences. By utilizing machine learning models and Spotify's API, the system identifies a user's music mood tendencies and provides insights similar to Spotify Wrapped.

Key Features

• Music Mood Prediction:

  • Analyzes users' liked songs and categorizes them into moods (Happy, Sad, Chill, Energetic).
  • Predicts the mood distribution of a user's music taste using machine learning models.

• Audio Feature Extraction:

  • Extracts key audio features from Spotify playlists and user-saved tracks using Spotipy.
  • Uses Essentia for in-depth rhythm and BPM analysis.

• Data Processing and Cleaning:

  • Filters out conflicting moods in duplicated song names.
  • Applies Z-score normalization and Standard Scaling to prepare data for model training.

• Machine Learning Classification:

  • Implements multiple classifiers (Random Forest, Decision Trees, SVM, Neural Networks, etc.).
  • Selects the best-performing model for mood prediction.

Technologies Used

• Programming Language: Python
• Audio Processing: Essentia, Pydub
• Machine Learning: Scikit-learn, NumPy, Pandas
• Spotify API: Spotipy
• Data Handling: JSON, CSV

How It Works

1. Data Collection:

   • Retrieves user's liked songs from Spotify.
   • Extracts features from mood-based Spotify playlists.

2. Preprocessing and Feature Engineering:

   • Drops irrelevant attributes and normalizes data.
   • Handles duplicates and conflicting labels.

3. Model Training & Evaluation:

   • Trains multiple classifiers and selects the best model using cross-validation.

4. Music Mood Prediction:

   • Predicts the mood distribution of a user's listening habits.
   • Outputs a percentage breakdown of each mood category.

Setup Instructions

1. Clone the repository:

   git clone https://github.com/ceasarattar/MusicAnalysis.git
   cd MusicAnalysis

2. Set up Spotify API credentials in library.py and spotify_mood_playlist.py.
3. Run the script to analyze music preferences:

   python modeling.py