BusyBody: Creating and Fielding Personalized Models of the Cost of Interruption

Eric Horvitz, Paul Koch, Johnson Apacible

Adaptive Systems & Interaction
Microsoft Research
Redmond, WA 98052

Access pdf.


Interest has been growing in opportunities to build and deploy statistical models that can infer a computer user’s current interruptability from computer activity and relevant contextual information. We describe a system that intermittently asks users to assess their perceived interruptability during a training phase and that builds decision-theoretic models with the ability to predict the cost of interrupting the user. The models are used at run-time to compute the expected cost of interruptions, providing a mediator for incoming notifications, based on a consideration of a user’s current and recent history of computer activity, meeting status, location, time of day, and whether a conversation is detected.

Keywords: Models of attention, multi-tasking, notifications, cost of interruptions, user modeling

In: Proceedings of CSCW, Conference on Computer Supported Cooperative Work, Chicago, Illinois, November 2004. ACM Press.

Related Background

Author Email: horvitz@microsoft.com

Back to Eric Horvitz's home page.