- Used a small dataset of MovieLens (100,000 ratings applied to 9,000 movies by 600 users) and the web-scraped movie posters from OMDb API
- Implemented the following recommender methods:
- Simple Recommender (recommend the most popular movies)
- Non-Negative Matrix Factorization (NMF)
- Collaborative Filtering
- Built a movie recommender app with Flask - user can select favorites movies. When user rate selected movies, 5 movies based on the NMF algorithm are recommended
- Clone the git repository:
git clone https://github.com/miladbehrooz/Movie_Recommender.git - Get API KEY from OMDb API and copy it to
flask-app/credentials.py - Install the requirements:
pip install requirements.txt - Run web app locally:
python flask-app/app.py