| Summary: | Multi-tasking is a common activity for computer users. Many recent approaches
to help support a user in multi-tasking require the user to indicate
the start (and at least implicitly) end points of tasks manually. Although
there has been some work aimed at inferring the boundaries of a user's
tasks, it is not yet robust enough to replace the manual approach. Unfortunately
with the manual approach, a user can sometimes forget to identify
a task boundary, leading to erroneous information being associated with a
task or appropriate information being missed. These problems degrade the
effectiveness of the multi-tasking support. In this thesis, we describe two interfaces
we designed to support task boundary identification. One interface
stresses the use of episodic memory for recalling the boundary of a task; the
other stresses the use of semantic memory. We investigate these interfaces
in the context of software development. We report on an exploratory study
of the use of these two interfaces by twelve programmers. We found that the
programmers determined task boundaries more accurately with the episodic
memory-based interface and that this interface was also strongly preferred. === Science, Faculty of === Computer Science, Department of === Graduate
|
|---|
