🎵 Instant Music Recommender A machine-learning-powered web application designed to suggest songs based on your current favorites. This project utilizes Natural Language Processing (NLP) to calculate similarities between tracks and provides a seamless, interactive interface for music discovery.

🚀 Project Overview This application is more than just a list; it is a recommendation engine built on data-driven insights. By analyzing song features and lyrics, it identifies tracks that "vibe" together using mathematical similarity scores.

Frontend: Streamlit (Python-based web framework)

Backend: Python

Machine Learning: TF-IDF Vectorization & Cosine Similarity

Dataset: Spotify Million Song Dataset

🧠 How It Works The core logic follows a standard Data Science pipeline to ensure accuracy and speed:

Vectorization: Song text and metadata are converted into numerical vectors using TF-IDF (Term Frequency-Inverse Document Frequency).

Similarity Matrix: The system employs Cosine Similarity to calculate the distance between these vectors. A smaller distance (closer to 1.0) indicates a higher similarity between songs.

Real-time Retrieval: Upon selecting a song, the app queries pre-computed .pkl models to instantly return the top 5 most relevant matches.