🎧 Bollywood Music Recommendation System 🇮🇳 An interactive Music Recommendation Web App built using Machine Learning and Streamlit that suggests Bollywood songs based on audio features, mood, and user preferences. The project also includes an AI-powered chatbot for conversational recommendations.

🚀 Features 🎵 Content-Based Recommendation System using cosine similarity 🎭 Mood-Based Filtering (Happy, Sad, Energetic) 🔍 Search Functionality for quick song selection 🤖 AI Chatbot Assistant for natural language queries 🎨 Modern UI (Spotify-inspired dark theme using Streamlit) ⚡ Optimized Performance with dataset sampling & caching

🧠 How It Works Extracts audio features like: Danceability Energy Tempo Loudness Valence Computes similarity between songs using: Cosine Similarity Provides recommendations based on: Selected song Mood filters Chatbot queries

📊 Dataset Based on a large Spotify dataset (~114K songs) Filtered to Bollywood artists for better relevance Reduced to ~1000 samples for performance optimization

🛠️ Tech Stack Python Pandas Scikit-learn Streamlit

🎯 Usage Select a Bollywood song from the dropdown Apply mood filters (optional) Click "Recommend Songs" Or use the 🤖 chatbot for smart suggestions

🤖 Chatbot Capabilities “Suggest sad Bollywood songs” “Songs by Arijit Singh” “Give me party songs”

💡 Key Highlights Solves real-world recommendation problem Handles large dataset optimization Implements interactive UI + chatbot