MAX (A Meta-reasoning Architecture for 'X') is a cognitive architecture, developed originally (1991) by Daniel R. Kuokka (under J. Carbonell) at Carnegie Mellon University, Pittsburgh, PA, USA.
MAX may be considered a "de-evolution" of Prodigy. The underlying logical support for modules is present in both, but MAX scales the rest of Prodigy's architecture down considerably.
D. R. Kuokka | MAX (Meta-reasoning Architecture "X") | See also The Prodigy Homepage | J. Carbonell’s Website
Когнитивная архитектура MAX [Kuokka, 1991] разработана в 1991 г. Даниэлом Куокка (Daniel R. Kuokka) под руководством Х. Карбонелла (Jaime Carbonell) в Университете Карнеги-Меллона (Carnegie Mellon University), Питтсбург, Пенсильвания, США во время совместной работы над архитектурой Prodigy.
В настоящее время Даниэл Куокка разрабатывает программное обеспечение в коммерческом секторе.
Description of MAX Architecture
MAX stands for Meta-Reasoning Architecture for "X", where "X" is an unbound variable.
MAX was designed to integrate reasoning, execution, and learning into an intelligent agent.
MAX was specifically designed to incorporate meta-reasoning. All cognitive actions, such as planning and learning, are controlled by meta-level knowledge.
Six components of the architecture are described in detail:
Interpreter
Logic Frames
Behaviors
Monitors
Domain Knowledge
Behavior Stack
The knowledge base in MAX is central to the architecture.
All knowledge, including meta-knowledge, is represented homogeneously by l-frames (logic-frames).
Perceptions are translated into knowledge, the execution of the problem-solver ("interpreter") is controlled by knowledge, and actions are taken from knowledge.
Possible agents include autonomous robots or system controllers.
MAX seeks to achieve planning, information gathering, and action execution, all in real time.
Publications
Selected papers
Kuokka, D.R.,"MAX: A meta-reasoning architecture for 'X'," SIGART Bulletin 2, 1991, 93-97.
