Movie Recommender System

A content-based movie recommendation system built using Python.
The system recommends movies similar to a given movie based on metadata such as genres, cast, crew, and overview.

How It Works

1. Movie metadata is preprocessed and combined into a single feature set
2. Text data is vectorized
3. Cosine similarity is used to measure similarity between movies
4. The top 5 most similar movies are recommended
5. A Streamlit web app is used for interaction (runs locally)

Features

• Content-based filtering (no user data required)
• Cosine similarity for recommendations
• Interactive Streamlit interface
• Displays movie posters and ratings (TMDB API)
• Runs locally using Jupyter Notebook / Streamlit

Tech Stack

• Python
• Pandas
• Scikit-learn
• Streamlit
• Jupyter Notebook

How to Run (Locally)

1. Clone the repository
2. Install dependencies: pip install -r requirements.txt
3. Install the pkl files Run the last code in the notebook to generate the required files
4. Run the streamlit app streamlit run app.py

Author

Rishi Kumar Machine Learning & AI enthusiast Github: Rishiii57