Context The collaborative filtering algorithm to build a recommendation engine is used with the user-item interaction information/data. Here the item is a movie and the user is the viewer. Content The data-set contains UserID, movieID and rating for each movie (by the user). The rating is a compound score (from 0 to 5) here. These are the three pieces of data (user/customer, item/product and an affinity (between the user and item) score required in the algorithm of collaborative filtering. Acknowledgements Google Cloud documentation Inspiration This can be used for building a recommendation engine for movie-goers.