Real-time context activity scheduling for smart space

A task management system that automatically optimizes the schedule of a person by context-aware interleaving of activities can enable persons to be involved more efficiently in daily activities of interests. As a technical basis for such a task management system, we propose the activity broker architecture. The core of this architecture is formed by the activity scheduler that can effectively schedule personal activities in different contexts in a real-time fashion. As a first step towards realizing the overall activity broker architecture, we present our solution algorithms to the Real-Time Context Activity Scheduling problem. Furthermore, we conduct extensive experiments to evaluate and to compare the performance of the proposed algorithms. The results show that a person can obtain more productive schedules with the assistance of the activity scheduler.