Spotify Music Recommendation System

This project builds a music recommendation system using K-Means clustering to group songs with similar features and a distance-based recommendation algorithm to suggest songs based on a given track. The dataset consists of Spotify album and feature data from 2023.

Motivation

With millions of songs available on streaming platforms, finding similar tracks can be challenging. This project applies machine learning techniques to analyze song features and recommend similar tracks.

Features

• Data Preprocessing:

  • Merges Spotify album data with song feature data.
  • Handles missing values and normalizes numeric features.

• Clustering Songs:

  • Uses K-Means clustering to group songs into clusters based on their audio features.

• Song Recommendation System:

  • Implements a distance-based recommendation algorithm.
  • Finds and suggests similar songs based on a given track.

Getting Started

Prerequisites

Ensure you have the following installed:

• Python 3.x
• NumPy
• Pandas
• Matplotlib & Seaborn (for visualization)
• Scikit-learn (for clustering & scaling)
• tqdm (for progress bars)

Installation

1. Clone the Repository

git clone https://github.com/your-username/spotify-recommendation-system.git
cd spotify-recommendation-system

2. Install Dependencies

pip install -r requirements.txt

3. Run the Jupyter Notebook

jupyter notebook Spotify_rec.ipynb

How It Works

1. Data Preprocessing

• Loads Spotify album data (spotify-albums_data_2023.csv) and track feature data (spotify_features_data_2023.csv).
• Merges the datasets on the track_id column.
• Cleans missing values and normalizes numeric features using MinMaxScaler.

2. K-Means Clustering

• Applies K-Means clustering to group songs into 35 clusters based on their feature similarity.

3. Recommendation System

• Uses a custom distance function to find and suggest songs most similar to a given track.
• Example:

  recommendations = Spotify_rec(df)
  recommendations.recommend("Lover", 13)  # Recommends 13 similar songs

To-Do

• Improve clustering by experimenting with more algorithms like Expectation Maximization.
• Enhance the recommendation model using additional feature engineering.
• Deploy the recommendation system using a web app (Streamlit)

License

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