Discover your next favorite track with Groovy

Features • Screenshots • Tech Stack • Architecture • Installation • How It Works • License

🚀 Features

• Personal Music Library - Browse and play Billboard's top tracks
• AI-Powered Recommendations - Personalized song suggestions that improve as you listen
• User Authentication - Create an account and sign in to access your personalized content
• Cross-Device Sync - Your listening history and preferences follow you across devices
• Federated Learning - Your listening data is securely stored and processed
• Billboard Hot 100 Integration - Automatically fetch the latest chart-toppers
• YouTube Integration - Download high-quality audio from official music videos
• Beautiful UI/UX - Modern, responsive design with intuitive controls

📸 Screenshots

Groovy's login page

Groovy's main interface with album cover display and music player

Personalized recommendations improve as you listen to more music

🏗 Architecture

🎵 Groovy
│
├── 📂 assets/                 # Assets directory
│   ├── 📂 imgs/              # Album artwork images
│   ├── 📂 meta/              # Song metadata JSON files
│   └── 📂 music/             # MP3 audio files
│
├── 📄 Groovy.py               # Main application UI
├── 📄 login.py                # Authentication interface
├── 📄 firebase_config.py      # Firebase configuration
├── 📄 rec.py                  # Recommendation system
├── 📄 fetch_hot_100.py        # Billboard scraper
├── 📄 download_music.py       # YouTube downloader
├── 📄 clear_db_assets.py      # Utility to reset app
├── 📄 run_groovy.py           # Application launcher
├── 📄 .env.example            # Environment variables template
└── 📄 requirements.txt        # Dependencies

📦 Installation

1. Clone the repository

   git clone https://github.com/rahulrao9/Groovy.git
   cd groovy

2. Install dependencies

   pip install -r requirements.txt

3. Set up environment variables

   cp .env.example .env

   Edit the .env file with your Firebase credentials and other configuration

4. Fetch Billboard Hot 100 data

   python fetch_hot_100.py

5. Download music from YouTube

   python download_music.py

6. Launch the app

   python run_groovy.py

   or

   streamlit run Groovy.py

7. Open the app in your browser

   http://localhost:8501
   

🎮 How It Works

Federated Learning Music Recommendation System

Groovy uses a sophisticated recommendation engine that combines multiple approaches:

1. Content-Based Filtering

   • Analyzes artists, song features, and tags
   • Creates embeddings to find similar music

2. Play Count Analysis

   • Tracks which songs you play most frequently
   • Builds an understanding of your explicit preferences

3. Artist Affinity Modeling

   • Calculates your preference for specific artists
   • Recommends new songs from artists you enjoy

4. Genre/Tag Preference Learning

   • Identifies music genres you prefer
   • Suggests similar music in your favorite styles

5. Hybrid Scoring

   • Combines all these signals with global popularity metrics
   • Produces personalized recommendations that improve over time

The recommendation model updates every 10 plays, adapting to your changing preferences.

User Authentication and Data Sync

Groovy uses Firebase Authentication to:

• Allow users to create accounts and sign in
• Securely store and sync listening history across devices
• Provide personalized recommendations that follow users wherever they go

Data Privacy

User data is securely stored in Firebase and processed according to best practices. Authentication ensures that your listening habits are private and only accessible to you.

🧠 Machine Learning Details

The recommendation system works by:

1. Preprocessing

   • Artist name normalization and splitting
   • Tag extraction and processing
   • Numerical feature normalization

2. Vectorization

   • Transforms musical attributes into mathematical spaces
   • Calculates similarities between songs

3. Multi-factor Scoring

   recommendation_score = (
       count_normalized * 0.35 +          # Play count
       engagement_ratio * 0.1 +           # How engaging the content is
       artist_affinity_score * 0.15 +     # Artist preference 
       tag_affinity_score * 0.15 +        # Genre/tag preference
       collaborative_score * 0.15 +       # Similar users' preferences
       popularity_score * 0.1             # Global popularity
   ) * (1 + recency_factor * 0.5)         # Recency boost

📋 Todo List & Future Improvements

• User accounts and profiles
• Multi-device sync
• Social sharing features
• Advanced visualization of listening habits
• Voice-activated controls
• Expanded music metadata

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Made with ❤️ by Rahul Rao

GitHub • Kaggle • LinkedIn