This paper describes the Prolog-Like Agent Specification & Tutor Implementation Calculus (PLASTIC) framework. PLASTIC is an intelligent tutor authoring system in which instructional agents based on models of human decision making, communication, action, and unit task execution are specified and deployed. To help non-programmers develop agent-based intelligent tutoring systems, PLASTIC includes GUI-based agent specification and performance visualization tools. PLASTIC agents provide instructional scaffolding by offering helpful hints that highlight task-relevant aspects of problems, help trainees think about tasks appropriately, guide trainees through unfamiliar problems, and encourage new strategy formation. To demonstrate how PLASTIC is used to implement agent-based instructional systems, a tutor incorporated into an Anti-Air Warfare Coordinator (AAWC) scenariobased training system is discussed. The technology through which PLASTIC tracks AAWC trainee actions and produces instructional scaffolding is discussed. The instructional impact of the PLASTIC tutor incorporated into the AAWC scenario-based trainer is illustrated with a discussion of training effectiveness data.
[1]
John R. Anderson,et al.
Cognitive Tutors: Lessons Learned
,
1995
.
[2]
Richard Reviewer-Granger.
Unified Theories of Cognition
,
1991,
Journal of Cognitive Neuroscience.
[3]
Kendall Scott,et al.
UML distilled - a brief guide to the Standard Object Modeling Language (2. ed.)
,
2000,
notThenot Addison-Wesley object technology series.
[4]
John R. Anderson,et al.
Rules of the Mind
,
1993
.
[5]
Bernhard Bauer,et al.
Extending UML for agents
,
2000
.
[6]
Allen Newell,et al.
The psychology of human-computer interaction
,
1983
.
[7]
John R. Anderson,et al.
Spanning seven orders of magnitude: a challenge for cognitive modeling
,
2002,
Cogn. Sci..
[8]
C. Heinze,et al.
Specifying Requirements in a Multi-Agent System with Use Cases
,
2000
.