Dataset: User side acquisition of People-Centric Sensing in the Internet-of-Things

Abstract

We envision a world where people-centric sensing and personalized services can be achieved without centralized data collection and processing or the reliance on video-based surveillance. To help assess the feasibility of building context-aware applications while allowing users to fully control their potentially sensitive data, we studied the feasibility of people-centric sensing using only user-side data acquisition. Using only off-the-shelf, energy restricted sensor kits in a smart environment together with an energy-efficient message exchange scheme implemented on top of Bluetooth Low Energy, the collected dataset provides insight on a continuous cyber-physical view from users' individual perspectives. The availability of the dataset encourages further studies of users' activities, for instance to perform distributed inference on users' social interactions and activity trajectories.