A roadmap for privacy preserving tourist recommendation system

Users’ privacy is one of the main concerns of users who use recommender systems in general and tourist recommender systems in particular, due to the fact that they must share personal information (like preferences and location) with the system in exchange for recommendations. The personal information collected by the system is used for creating user models used for personalization of recommendations, but may be used and / or shared or sold to 3rd parties. Still, when considering content-based recommender systems, the situation may be different if the user’s model is built, maintained and stored locally on the user’s device/personal cloud. The paper presents a simple yet effective privacy preserving content-based recommender system architecture that uses a hypercube-based model for representing user preferences.