Movie Recommender System

This project is a Movie Recommender System built with Streamlit, using cosine similarity to recommend movies based on a selected title. It leverages movie metadata and the OMDb API to fetch movie posters, providing an engaging, interactive experience for users to discover similar movies.

Features

• Movie Recommendation: Based on a selected movie, the system provides recommendations for five similar movies.
• Poster Display: Movie posters are fetched using the OMDb API, offering a visual appeal.
• Interactive Interface: The application is deployed with Streamlit for a user-friendly, interactive experience.

Setup Instructions

1. Clone the repository

git clone https://github.com/yourusername/movie-recommender-system.git
cd movie-recommender-system

2. Install required libraries

Install the necessary packages listed in the requirements.txt file:

pip install -r requirements.txt

3. Obtain OMDb API Key

1. Sign up at OMDb API to get an API key.
2. Replace api_key=YOUR_API_KEY in the fetch_poster function with your OMDb API key.

4. Run the Streamlit app

streamlit run app.py

File Structure

• app.py: Main file containing the Streamlit app code.
• movies.pkl: A pickled file containing movie metadata.
• similarity.pkl: A pickled file with the precomputed similarity matrix.
• requirements.txt: Lists all required Python libraries for the project.

Project Details

The application is built around content-based filtering:

• Movie Metadata: The system uses metadata about each movie to compute a similarity matrix using cosine similarity.
• Recommendation Function: The recommend function retrieves the top five similar movies for a selected movie.
• Poster Fetching: The fetch_poster function uses the OMDb API to fetch poster images for each recommended movie.

Usage

1. Select a Movie: Choose a movie title from the dropdown menu.
2. Show Recommendation: Click on "Show Recommendation" to view five similar movie recommendations with posters.
3. View Recommendations: Recommended movies with their posters will be displayed in a 5-column layout.

Example

Selecting "Inception" from the dropdown and clicking "Show Recommendation" will display five movies similar to "Inception," with their posters.

Selecting "Spider-Man 3" from the dropdown and clicking "Show Recommendation" will display five movies similar to "Inception," with their posters.

Future Improvements

• Enhanced Recommendation Algorithm : Experimenting with other recommendation algorithms for improved recommendations.
• Expanded Movie Metadata: Incorporating genres, actors, and directors to improve similarity calculations.
• User Ratings Integration: Allowing users to rate movies, potentially enhancing personalized recommendations.