RADAR Personal Cognitive Assistant

Asim Smailagic, Human Computer Interaction
Dan Siewiorek, Human Computer interaction

The system monitors user inputs and attention focus, and classifies user activity patterns. This should help the user to stay task-focused. It also helps to minimizes efficiency loss due to cognitive context swapping while maintaining user responsiveness. Building on our activity identification technology, the system will be able to launch tasks, report on accomplishments, and orchestrate user task context swaps when agent-initiated activities have assembled new information requiring user input to proceed. By being able to identify tasks and task sequences, the system can help a user, in particular a novice, to more effectively use it; for example, help them when they are confused and cycling. The system also provides reminders to users to perform important functions as deadlines approach. It receives input on task priority plus degree of completion and focuses the user’s attention on high-value activities. An example project is to compare the effectiveness of raising task priorities vs. pop-up reminder messages. Applications can include: processing a large number of emails for busy managers, conference / meeting resource scheduling under a defined set of constraints, posting briefings, etc.