View article

We describe the design, functionality, and early-stage evaluation of TaskMinder, a to-do list management application. TaskMinder builds in rich environmental context, including location and user activity, and uses a machine learning algorithm to compare against a user’s history to produce an automatic ranking of to-do list items. The application also can suggest tasks to be accomplished given short term time constraints. We present key aspects of the TaskMinder design, discuss its implementation and evaluation, and suggest further work in the field of adaptive personal information agents.