The Soar Cognitive Architecture

In development for thirty years, Soar is a general cognitive architecture thatintegrates knowledge-intensive reasoning, reactive execution, hierarchical reasoning, planning, andlearning from experience, with the goal of creating a general computational system that has the samecognitive abilities as humans. In contrast, most AI systems are designed to solve only one type ofproblem, such as playing chess, searching the Internet, or scheduling aircraft departures. Soar isboth a software system for agent development and a theory of what computational structures arenecessary to support human-level agents. Over the years, both software system and theory haveevolved. This book offers the definitive presentation of Soar from theoretical and practicalperspectives, providing comprehensive descriptions of fundamental aspects and new components. Thecurrent version of Soar features major extensions, adding reinforcement learning, semantic memory,episodic memory, mental imagery, and an appraisal-based model of emotion. This book describesdetails of Soar's component memories and processes and offers demonstrations of individualcomponents, components working in combination, and real-world applications. Beyond these functionalconsiderations, the book also proposes requirements for general cognitive architectures andexplicitly evaluates how well Soar meets those requirements.

John E. Laird is John L. Tishman Professor of Engineering in the Computer Science Division at the University of Michigan.