A group preference-based privacy-preserving POI recommender system

Ubiquitous smartphones with their built-in location services enable people to explore various points-of-interest (POIs) through location-based apps, e.g., Yelp and Foursquare City Guide. With these apps, users can receive personalized recommendations on nearby places, e.g., restaurants and arcades, which not only saves them searching time, but also helps find POIs that are of interest to them. One issue with these apps and almost all existing recommender systems is that they require users to share their preference data with the service providers. This information, if not properly used, can leak users’ privacy. In this paper, we propose a group preference-based POI recommendation scheme which fuses matrix factorization and clustering techniques to provide quality recommendations without sacrificing users’ privacy.