codermillat/Music-Recommender-System

A comprehensive music recommendation system using collaborative filtering, content-based, and popularity-based approaches.

Music Recommender System

A comprehensive music recommendation system using collaborative filtering, content-based, and popularity-based approaches, optimized for Google Colab.

Features

1. Popularity-Based Recommender

   • Recommends most popular songs based on play counts
   • Simple but effective baseline system

2. Collaborative Filtering

   • User-based approach using KNN with Means
   • Cosine similarity metric
   • Optimized for sparse matrices

3. Content-Based Recommender

   • Uses song metadata (artist, title)
   • TF-IDF vectorization
   • Efficient similarity computation

Usage in Google Colab

1. Clone the repository:

!git clone https://github.com/codermillat/Music-Recommender-System.git

2. Install dependencies:

!pip install -r requirements.txt

3. Run the Streamlit app:

!streamlit run app.py

Dataset Structure

The dataset should be placed in the /data directory:

• kaggle_visible_evaluation_triplets.txt: User listening history
• unique_tracks.txt: Song metadata

Requirements

See requirements.txt for detailed dependencies.