Search

The Amulet user interface development environment makes it easier for programmers to create highly interactive, graphical user interface software for Unix, Windows and the Macintosh. Amulet uses new models for objects, constraints, animation, input, output, commands, and undo. The object system is a prototype instance model in which there is no distinction between classes and instances or between methods and data. The constraint system allows any value of any object to be computed by arbitrary code and supports multiple constraint solvers. Animations can be attached to existing objects with a single line of code. Input from the user is handled by "interactor" objects which support reuse of behavior objects. The output model provides a declarative definition of the graphics and supports automatic refresh. Command objects encapsulate all of the information needed about operations, including support for various ways to undo them. A key feature of the Amulet design is that all graphical objects and behaviors of those objects are explicitly represented at run time, so the system can provide a number of high level built-in functions, including automatic display and editing of objects, and external analysis and control of interfaces. Amulet integrates these capabilities in a flexible and effective manner.

RADAR is a multiagent system with a mixed-initiative user interface designed to help office workers cope with email overload. RADAR agents observe experts to learn models of their strategies and then use the models to assist other people who are working on similar tasks. The agents' assistance helps a person to transition from the normal email-centric workflow to a more efficient task-centric workflow. The Email Classifier learns to identify tasks contained within emails and then inspects new emails for similar tasks. A novel task-management user interface displays the found tasks in a to-do list, which has integrated support for performing the tasks. The Multitask Coordination Assistant learns a model of the order in which experts perform tasks and then suggests a schedule to other people who are working on similar tasks. A novel Progress Bar displays the suggested schedule of incomplete tasks as well as the completed tasks. A large evaluation demonstrated that novice users confronted with an email overload test performed significantly better (a 37% better overall score with a factor of four fewer errors) when assisted by the RADAR agents.

The RADAR project developed a large multi-agent system with a mixed-initiative user interface designed to help office workers cope with email overload. Most RADAR agents observe experts performing tasks and then assist other users who are performing similar tasks. The interaction design for RADAR focused on developing user interfaces that allowed the intelligent functionality to improve the user’s workflow without frustrating the user when the system’s suggestions were either unhelpful or simply incorrect. For example with regards to autonomy, the RADAR agents were allowed much flexibility in selecting ways to assist the user, but were restricted from taking actions that would be visible to other people. This policy ensured that the user remained in control and mitigated the negative effects of mistakes. A large evaluation of RADAR demonstrated that novice users confronted with an email overload test performed significantly better, achieving a 37% better overall score when assisted by RADAR. The evaluation showed that AI technologies can help users accomplish their goals.

Jadeite is a new Javadoc-like API documentation system that takes advantage of multiple users' aggregate experience to reduce difficulties that programmers have learning new APIs. Previous studies have shown that programmers often guessed that certain classes or methods should exist, and looked for these in the API. Jadeite's “placeholders” let users add new “pretend” classes or methods that are displayed in the actual API documentation, and can be annotated with the appropriate APIs to use instead. Since studies showed that programmers had difficulty finding the right classes from long lists in documentation, Jadeite takes advantage of usage statistics to display commonly used classes more prominently. Programmers had difficulty discovering how to instantiate objects, so Jadeite uses a large corpus of sample code to automatically the most common ways to construct an instance of any given class. An evaluation showed that programmers were about three times faster at performing common tasks with Jadeite than with standard Javadoc.

Calendar scheduling is a difficult task for people who have overbooked calendars with many constraints. Currently, calendar applications do not allow users to specify scheduling constraints such as how preferable a free time is for scheduling a new meeting or to what extent an existing meeting can be rescheduled. This paper introduces the "availability bar," an interaction and visualization technique for complex calendar scheduling constraints. Availability bars, embedded in calendar applications, can help users who manually schedule meetings. Availability bars can also mediate communication with calendar scheduling agents that gather availability constraints, search for times that satisfy the constraints, and negotiate with invitees when no satisfactory time is found for the constraints.

The RhaiCAL system provides novel visualizations and interaction techniques for interacting with an intelligent agent, with an emphasis on calendar scheduling. After an agent interprets natural language containing meeting information, a user can easily correct mistakes using RhaiCAL's clarification dialogs, which provide the agent with feedback to improve its performance. When an agent proposes actions to take on the user's behalf, it can ask the user to confirm them. RhaiCAL uses novel visualizations to present the proposal to the user and allow them to modify the proposal, and informs the agent of the user's actions in a manner that supports long-term learning of the user's preferences. We have designed a high-level XML-based language that allows an agent to express its questions and proposed actions without mentioning user interface details, and that enables RhaiCAL to generate high-quality user interfaces.