Movie-Recommender

A web-based movie recommender engine, recommends movies to the user in one of two algorithms:

• Non-negative matrix factorization
• Collaborative Filtering: User-based Filtering recommendation

The dataset used is : MovieLens dataset https://grouplens.org/datasets/movielens/

Usage:

• clone the repository

• install the requirements.

• login to your postgres and create a database called: movies_db

  so you should have the following connection: connection = postgresql://localhost:5432/movies_db

• run data_preprocessing.py , so the data is read from csv files, processed and stored in movies_db

• you can choose in application.py between the two algorithms:

  • recommended_movies = recommend_NMF()
  • recommended_movies = user_based_filtering_recommend()

• run application.py and go to your browser and type: localhost/5000

the use has to input five movies and rate them:

and then the user gets his recommended movies:

Tech used:

• Python
• Flask
• HTML
• CSS
• PostgreSQL
• sqlalchemy
• Scikit-learn

License

• MIT license