Firstly, we read the ratings from a file. For this project, we can use the MovieLens 100k rating dataset from http://www.grouplens.org/node/73. The training set ratings are in a file called ua.base, while the movie item data is in u.item. On the other hand, ua.test contains the test set to evaluate our model. Since we will be using this dataset, we should acknowledge the GroupLens Research Project team at the University of Minnesota who wrote the following text:

F. Maxwell Harper and Joseph A. Konstan. 2015. The MovieLens Datasets: History and Context. ACM Transactions on Interactive Intelligent Systems (TiiS) 5, 4, Article 19 (December 2015), 19 pages. DOI: http://dx.doi.org/10.1145/2827872.

This dataset consists of 100,000 ratings of 1 to 5 from 943 users on 1,682 movies. Each user has rated at least 20 movies. It also contains simple demographic info about the users (age, gender, occupation, and zip code).