Students will design a piece of software that performs the following four
tasks:
1. Observation
Software monitors user behavior and external events, watching for the correct
moment to take an action.
2. Learning
Software monitors user behavior and draws inferences such as a users preference
or habit. Software that monitors events to learn general patterns.
3. Dialogue
Software that engages users in communication. Generally communication facilitates
clarification or acknowledgement. Dialogue generally takes three forms: (i)
acknowledgement—confirms an action has happened. (ii) turn taking—seeks
clarification or confirmation following user request. (iii) monitoring—proactively
seeks clarification or confirmation based on observation activities.
4. Autonomous Action
Software that takes action on its own whether preprogrammed or from learning.
Deliverables
1. Short video demonstration of human agent interaction in context.
2. Presentation outlining user needs
3. Web-based process book
Alla
Moda - Grace Chang and Jeff Park
An agent that helps shoppers pick clothes within a store that complement their
existing wardrobe and are within a certain price range.
BioPlanner - Ben Ledewitz
, Lindsey Cortese
An agent that learns to correlate your activities with certain emotional states
(especially stress) to let you know how you will be emotionally affected by
the day's schedule.
Wardrobe Agent - Nick Canilang, Lily Cho, Roald Munoz
An agent that prepares clothing for one or more
days based on the type of activity, weather, and other factors.
Smart Kitchen - Becky Katterson,
Tim Kelly, and Andrea Wagner
An agent for the kitchen that could help one cook with more safety, assurance,
and success.
Sharing Space Made Smart - Agata Adamowicz,
Engly Chang, and Janice Lau
An agent for the Smart Home for college students, designed to improve the
living environment by making it easier to share the living space with multiple
people.
Interaction and Visual Interface | 51-421 | Fall 2003
|