CSAIL Publications and Digital Archive header
bullet Technical Reports bullet Work Products bullet Research Abstracts bullet Historical Collections bullet

link to publications.csail.mit.edu link to www.csail.mit.edu horizontal line


Research Abstracts - 2006
horizontal line

horizontal line

vertical line
vertical line

Gradual Awareness Notification for Desktop Systems

Tom Wilson & Robert Miller


Current notification schemes for desktops are, for the most part, insensitive to the user. They interrupt users at precisely the wrong moment, leading to frustration, annoyance, and lost productivity. As more and more applications vie for attention on desktop systems, traditional notification systems fail even more dramatically. In particular, with mobile and ubquitous systems growing in importance every day, current notification techniques will be entirely unacceptable. Clearly, a more intelligent and respectful method of informing users is needed.

Much work has been done on the subject of interruptibility. Most results indicate that there are distinct costs for interruption at non-optimal moments. Specifically, Bailey and Iqbal show that mental workload significantly decreases at task boundaries, and that interrupting users at these task breaks can substantially reduce the ill-effects of interruption. However, the challenge of isolating and identifying these boundaries remains unsolved. Several systems have been proposed to implement ideal interuption. For the most part, though, these systems are complex, inefficient, incomplete, and may require training to adapt to practices of particular users or tasks.

Rather than attempt to predict opportune moments, we hypothesize that by using a clever notification technique which takes advantage of change blindness, systems can better achieve interruption at task boundaries in an efficient, simple, and light-weight manner.


We have developed a model for gradual awareness notification. Specifically, we focus on slow growth windows. Traditional desktop notifications usually work by popping up windows to inform users. This is problematic because it causes a large amount of change directly in the users field of focus, leading to a high degree of distraction.

In contrast, our slow growth system gradually grows notifications from the side of the screen. This achieves several benefits. Firstly, because the window is gradually encroaching, the rate of change of the screen is substantially reduced. When focusing on a task, people suffer from an effect known as "change blindness," where gradual changes in surroundings go almost entirely unnoticed. Secondly, as the window gets bigger, the user becomes more likely to notice it. When users are focusing on a particular task, they enter a state of tunnel vision. At task breaks, their effective field of vision tends to widen, allowing them to take in a bigger field surrounding their task. At that moment, they are likely to notice the notification.

a sample slow growth window

Thus, by gradually growing the windows, we hypothesize that users will tend to notice the notifications at precisely those moments when they are most ready to be interrupted. The major benefit of this system is that we no longer need to have any information about the user's task or the user themselves.


Currently, we have developed prototypes for both a lab and a field study. In the coming months, we plan on conducting user studies in order to quantify the benefits of gradual awareness notifications. In the field study, users will run a small program which randomly displays slow growth notifications. In the lab study, we ask users to complete a task while periodically interrupting them with different types of notifications. We hope that the results from these studies will provide better insight into properties of interruptibility.


[1] Bailey, B.P and S.T. Iqbal. Leveraging Changes in Mental Workload during Task Execution to Mitigate Effects of Interruption. Technical Report UIUCDCS-R-2005-2623, University of Illinois, August, 2005. (Length: 47 pages).

[2] Horvitz, E., Koch, P., and Apacible, J. 2004. BusyBody: creating and fielding personalized models of the cost of interruption. In Proceedings of the 2004 ACM Conference on Computer Supported Cooperative Work (Chicago, Illinois, USA, November 06 - 10, 2004). CSCW '04

[3] Iqbal, S.T. MeWS-IT: A Mental Workload based System for Interruption Timing. Proceedings of the ACM Symposium on User Interface Software and Technology, Doctoral Symposium, October, 2005.

[4] Simons, D. J., & Chabris, C. F. (1999). Gorillas in our midst: Sustained inattentional blindness for dynamic events. Perception, 28, 1059-1074.


vertical line
vertical line
horizontal line

MIT logo Computer Science and Artificial Intelligence Laboratory (CSAIL)
The Stata Center, Building 32 - 32 Vassar Street - Cambridge, MA 02139 - USA
tel:+1-617-253-0073 - publications@csail.mit.edu