This project deals with generating movie recommendations by predicting the likability of a movie by a user based on the response of other similar users, where this similarity is calculated based on the movie ratings left by other users. The system is based on Collaborative Filtering (CF) with a modification, where we embed our users and movies into a low dimensional space to learn these similarities among users and movies.
Dataset: MovieLens 20M dataset made available by GroupLens. This dataset contains 20 million ratings and 465, 000 tag applications applied to 27, 000 movies by 138, 000 users.