Reasoning About Rational AgentseBook - 2000
One goal of modern computer science is to engineer computer programs that can act asautonomous, rational agents; software that can independently make good decisions about what actionsto perform on our behalf and execute those actions. Applications range from small programs thatintelligently search the Web buying and selling goods via electronic commerce, to autonomous spaceprobes. This book focuses on the belief-desire-intention (BDI) model of rational agents, whichrecognizes the primacy of beliefs, desires, and intentions in rational action. The BDI model hasthree distinct strengths: an underlying philosophy based on practical reasoning in humans, asoftware architecture that is implementable in real systems, and a family of logics that support aformal theory of rational agency.The book introduces a BDI logic called LORA (Logic of RationalAgents). In addition to the BDI component, LORA contains a temporal component, which allows one torepresent the dynamics of how agents and their environments change over time, and an actioncomponent, which allows one to represent the actions that agents perform and the effects of theactions. The book shows how LORA can be used to capture many components of a theory of rationalagency, including such notions as communication and cooperation.
This book focuses on the belief-desire-intention (BDI) model of rational agents,which recognizes the primacy of beliefs, desires, and intentions in rational action.