Patent attributes
This disclosure is directed to a system and method that generates personalized recommendations of media items for customers based on a customer's preference for familiar or novel media items. A customer can be associated with user data that can include data such as a customer's purchase data, shopping data, etc. The user data can be used to determine media item data that is associated with the customer. Based on the customer's media item data, a familiarity score can be generated that indicates a media consumption preference of the customer. Using the familiarity score, a recommendation identifying a media item can be generated for the customer.