|
|
|
Mark Foltz |
|
mfoltz@ai.mit.edu |
|
April 17, 2002 |
|
Intelligent Room HCI Group |
|
|
|
|
|
|
|
|
|
|
A central aspect of our mental life |
|
Visual and speech perception, discourse, … |
|
Competition between salience (stimulus-driven)
and intention (goal-driven) |
|
|
|
|
Bandwidth mismatch |
|
|
|
No Moore’s Law for humans |
|
|
|
Pervasive = more info sources, more competition
for attention |
|
|
|
A research opportunity |
|
|
|
|
“Query by Attention,” Foltz and Davis 2001 |
|
|
|
“Notification, Disruption, and Memory,” Cutrell,
Czerwinski, Horvitz 2001 |
|
|
|
Discussion/Brainstorming: Implications for E21 |
|
|
|
|
|
|
|
|
|
|
User controls attention, instead of manipulating
an interface |
|
|
|
Rapid adjustment of query parameters |
|
|
|
Immediate feedback |
|
|
|
Can lead to more complete exploration of info
space |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Interrupt or not? |
|
If so, when (now or later)? |
|
|
|
|
|
|
|
|
Interrupt or not? |
|
If so, when (now or later)? |
|
And whom (individual, group, a delegate)? |
|
|
|
|
|
|
|
|
Interrupt or not? |
|
If so, when (now or later)? |
|
And whom (individual, group, a delegate)? |
|
|
|
And, how can we remember where we were? |
|
|
|
|
|
|
|
|
|
|
|
|
|
Task |
|
Subjects search a list of book titles |
|
Given either a title or a description (gist) |
|
While being interrupted |
|
|
|
Conditions (2 x 2 x 2) |
|
Title vs. gist search |
|
Cursor vs. no cursor |
|
Interruptions vs. no interruptions |
|
|
|
16 subjects |
|
|
|
|
|
Gist search has a high reaction time |
|
|
|
Title search has a high recovery time |
|
|
|
Interruptions are more disruptive early in the
primary task |
|
More reminders requested |
|
|
|
|
|
|
|
Don’t interrupt early in the primary task. |
|
Focus the interruption on whom it affects. |
|
Assist swap-in. |
|
Record the context before the interruption. |
|
Create a virtual secretary. |
|
E.g., knows the priority of the current meeting. |
|
Group interruptions by topic. |
|
|
|
Notes