AI Seminar
    -------------------------------
    Tuesday, December 2nd, 2003
    4:00 pm - 5:30 pm
    175 ATL (Large Conference Room)

    "EXAMPLE-DRIVEN DIAGRAMMATIC TOOLS FOR RAPID KNOWLEDGE ACQUISITION"

    John Laird
    Department of Electrical Engineering and Computer Science
    University of Michigan




----------------------------------
The last ten years has seen a revolution in the complexity and realism of
computer models of complex human behavior. However, the cost of developing
realistic these models continues to increase as much of the detailed and
complex knowledge must be manually encoded to produce realistic behavior.
The focus of this project is on reducing the cost of acquiring, validating
and maintaining the knowledge used in realistic human behavior models.  Our
approach is to develop tools that allow subject matter experts (SMEs) to
visually specify behavior using abstract scenarios represented as diagrams.
The SME can graphically describe the conditions under which actions and
goals should be pursued, together with the associated reasons for those
decisions.  The system, guided by the expert's choices, analyzes and
automatically generalizes from the example scenarios, alerting the SME to
inconsistencies and missing knowledge.  The system incrementally generates
an executable models whose behavior the SME can view and modify during
development.  By moving the language of discourse from symbolic programming
languages to annotated diagrams, the SMEs specify knowledge directly
with-out requiring the intervention of a knowledge engineer to translate
between the representations.

This work was done in conjunction with Doug Pearson of Three Penny Software.