Reinforcement Learning Based Smart Data Agent for Location Privacy

Abstract

A “privacy by design” approach is presented for location management on a smartphone that is based on the concept of “smart data”. Extensive use of location-based services has given rise to various privacy concerns such as processing and selling of personal information. The proposed work addresses these concerns by developing an intelligent agent that controls the release of the user’s information in accordance with its preferences, context, nature of situation. The proposed model uses reinforcement learning to provide an on-the-go smart data agent that learns the user’s privacy policy in a manner that is interactive and adaptive, enabling it to adjust itself to changes in user preferences over time. The agent aimed at mimicking the user, deciding when and where exact location information should be revealed and evolving over time to become the user’s single trusted virtual proxy in cyberspace, managing its interactions with various applications in different environments. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.