Learning and reasoning about interruption

Abstract

We present methods for inferring the cost of interrupting users based on multiple streams of events including information generated by interactions with computing devices, visual and acoustical analyses, and data drawn from online calendars. Following a review of prior work on techniques for deliberating about the cost of interruption associated with notifications, we introduce methods for learning models from data that can be used to compute the expected cost of interruption for a user. We describe the Interruption Workbench, a set of event-capture and modeling tools. Finally, we review experiments that characterize the accuracy of the models for predicting interruption cost and discuss research directions.

Interruptions in the workplace: A case study to reduce their effects

“…By far the most popular form of interruption that computer users encounter on a day-to-day basis is email notifications and instant messenger client pop-up messages (Gievska, Lindeman, & Sibert, 2005;Horvitz & Apacible, 2003;Iqbal & Bailey, 2007). For the content of the interruption to be presented to the user, the user must first acknowledge the interruption-in other words the user must be attentive to the interruption notification.…”

Predicting time-sharing in mobile interaction

“…Third, echoing points made by others before us (Fogarty et al 2005;Horvitz and Apacible 2003), we believe that there is no single measure of "available time" that would support all adaptations. We will therefore use the following four variables for describing complementary aspects of time-sharing within the predicted time window:…”

“…Some of the most elaborate models to date for adapting to the attentional states of the user have been constructed at Microsoft Research (Horvitz 1999b;Horvitz et al 1999Horvitz and Apacible 2003). The work attempts to provide a foundation for both enhancing existing applications and creating new kinds of applications based on mixed-initiative computing.…”

“…To critically assess the feasibility of such adaptations, we present a wizard-of-oz feasibility study (Hudson et al 2003;Fogarty et al 2005) looking at automatic prediction of time-sharing based on sensors of varying degrees of sophistication. Rather than just pooling results from empirical work, we contribute to the field by examining the possibility of using predictive models for the development of real-world computing and communication applications (see Horvitz and Apacible 2003). Our analysis and modeling focuses on the sharing of time in mobile interaction, one of the increasingly more important domains of user modeling and adaptation (Kobsa 2001).…”

Handling conflicts of context-aware reminding system in sensorised home

“…Cyberminder [3] and ComMotion [9] are two examples of context-aware systems which attempt to build a context-aware infrastructure to support reminder applications. The issue of how and when to prompt subjects has also been examined in [2,6,7] from other research angles. 3 System design Figure 1 shows the architecture of the context-aware reminding system.…”