Movie Recommendation System

Overview

This machine learning project implements a content-based movie recommendation engine using TF-IDF vectorization and cosine similarity. Users can search for any movie and receive a list of semantically similar recommendations based on plot descriptions, genres, keywords, cast, and director. The model is served through a Streamlit web application with a clean, interactive UI.

Key Features

• TF-IDF vectorization combined with cosine similarity for content-based filtering
• Semantic search across plot, genre, keywords, cast, and director
• Top-N movie recommendations ranked by similarity score
• Interactive Streamlit web interface for real-time queries
• Pre-processed TMDB dataset with 5,000+ movies
• Scikit-Learn pipelines for reproducible model building