An interactive movie recommendation system that recommends the user top movies based on user's choices, using content-based filtering. Here's the dataset we collected from different sources: https://github.com/TheOneSSA/IMDb_Movies_Datasets
For deployment, we used Flask web framework and the Heroku platform. Link to the web application :- https://project-movie-recommender.herokuapp.com/ (some features aren't working properly, as Heroku has a 512 MB dyno limit under free subscription, but it will run properly on deploying on your local machine, with RAM >= 2 GB)
In the Movie Recommender.ipynb file the Data Preprocessing part has been done. The sparse matrices for each genre has also been saved in the folder - sparse_matrices.
The application is run from the app.py file.
Procfile and requirements.txt consists of all the requirements for deployment in Herolu platform.
Folder - datasets and sparse_matrices has genre-wise datasets + 1 master dataset of 75k+ movies and sparse matrices containing the Tf-idf scores of each, respectively.
Folder - templates has all the html files, for diiferent pages.
Folder - static has all the files for flask routing.