Year: 1994
Type: dissertation
Abstract: This work addresses three issues in intelligent tutoring systems. \nFirst, how to explain and illustrate the functioning of physical \ndevices. This capability is useful for understanding and using past designs. \nThe second issue is how to illustrate and explain the reasoning of an \nexpert designer. This capability is often required for supporting learning by \nobservation. Finally, might the same kinds of representations and \nprocesses support these two capabilities? This thesis explores how \nStructure-Behavior-Function models of physical devices and \nStructure-Behavior-Function meta models of an autonomous design system \nhelp to address these issues. The Structure-Behavior-Function model of \na device (or a system) explicitly specifies its structure, its \nfunction and its internal behaviors that map its structure into its \nfunction. This research has resulted in a system called Canah-chab \nthat provides an interactive graphical interface to an autonomous \ndesign system called Kritik2. It illustrates and explains both the \nphysical devices in Kritik2's memory and the processes by which it \ngenerates new design. In addition, Canah-chab enables a user to \nnavigate Kritik2's memory and access the stored designs.
Author: Nathalie Grué
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2501009142