Collaborative filtering algorithm to predict user ratings for products, based on previous ratings.
Here I am sharing a python implementation of a collaborative filtering algorithm used to solve the Netflix problem. We have a bunch of users that we do not know anything about them except the ratings that they gave to some of the films in our list. We want to use just this information to predict their possible ratings to other movies in the list that they haven't rated so far.
Most importantly, the jupyter notebook file collaborative_filtering_recommender.jpynb that guides you through the algorithm step by step. The dataset which is a small part of the original dataset also provided here.
As I mentioned, it is meant to be like a tutorial implementation and not an efficient implementation, but I've tried to make it efficient as long as it does not hurt the purpose of the project.
Yes, as long as you frame your data similar to the dataset used in this project. Look at the code or the description file in the netflix folder to know more about the data.
Just open the file by jupyter notebook and run all the cells.