In this repository you can find a dataset used for creating and evaluating recommender systems that have to deal with cold start problem.
There are three files:
movies.csv: contains data about the items (movies in this case).user_profiles.csv: contains data related to the user's characteristic and behaviour (this profile has been created based on the users' twitter profile).ratings.csv: contains all the ratings from the users for the movies.
The data has been extracted from Filmaffinity website and twitter.