Jump to the Architectural Overview Menu page.
Jump to the Home page.

An Overview of the PRODIGY Architecture

PRODIGY combines learning and planning to solve problems in particular domains. Planning in PRODIGY consists of finding a sequence of operators that can be applied to reach a specified goal. Several methods of learning are employed to fascilitate the generation of efficient plans. Knowledge gained through learning is added to the store of basic domain knowledge with which PRODIGY begins. This approach is intended to enable the system to become adept at solving problems in a given domains.

Characteristics



Sources and References


Jaime G. Carbonell, Craig A. Knoblock, Steven Minton. "Prodigy: An
Integrated Architecture for Planning and Learning" in Kurt VanLehn (Ed.),
 Architectures for Intelligence, pp. 241-278, Lawrence Erlbaum 
Associates, Hillsdale, NJ, 1991.


Carbonell, Jaime G., etal., "PRODIGY: An Integrated Architecture for
Planning and Learning",  SIGART Bulletin 2, 1991, pp. 38-42. 



The PRODIGY Manual is  available by ftp from reports.adm.cs.cmu.edu/usr/anon/1992/ as CMU-CS-92-150.ps.