Recommendations-based location privacy control

In this paper, we propose and investigate a user-centric device-cloud architecture for intelligently managing user data. The architecture allows users to keep their (private) data on their mobile devices and decide what to share with the service providers on the cloud, based on their individual privacy preferences, in order to get personalized services. Our architecture strives to help ease users' burden on managing privacy by giving automatic recommendations on how to configure their privacy profiles on devices. One focused contribution of this paper is that we instantiate this proposed general architecture to location-based service due to the privacy sensitivity of location data. We derive and validate our location-sharing recommendations using online user experiments. Our results show that the recommendations are accurate, and that they help users with the decisions involved in the privacy profile configuration process. Our results also demonstrate that the quality of personalized location-based services can be maintained even when the increased user privacy control leads to a situation where not all location data is shared with the service provider. These results lead the way to powerful location-based and other personalized services that improve user privacy.