Any recommender system problem can be posed both as a matrix factorization as well as regression problem. This is a beautiful project where I have tried to combine features from both to ultimately solve the problem.
The uploaded pdf file describes my research workflow. Please read it.
Heroku deployment URL: https://netflix-recommender-system-api.herokuapp.com
Please visit this site (deployed on AWS Elastic Beanstalk) (inactive now): http://netflixrecommendationsystem-env.eba-iat8iy3y.us-east-2.elasticbeanstalk.com/
Credits for front-end design of the webpage: https://github.com/aditya083-etce
- Clone the repository and open the command prompt in the clone directory.
- Install required packages
pip install -r requirements.txt - Run the 'application.py' file.
python application.py