An Agent Language with Destructive Assignment and Model-Theoretic Semantics

In this paper we present an agent language that combines agent functionality with an action theory and model-theoretic semantics. The language is based on abductive logic programming (ALP), but employs a simplified state-free syntax, with an operational semantics that uses destructive assignment to manipulate a database, which represents the current state of the environment. The language builds upon the ALP combination of logic programs, to represent an agent's beliefs, and integrity constraints, to represent the agent's goals. Logic programs are used to define macro-actions, intensional predicates, and plans to reduce goals to sub-goals including actions. Integrity constraints are used to represent reactive rules, which are triggered by the current state of the database and recent agent actions and external events. The execution of actions and the assimilation of observations generate a sequence of database states. In the case of the successful solution of all goals, this se...

• Type: Conference
• Year: 2010
• Where: CLIMA
• Authors: Robert A. Kowalski, Fariba Sadri
• Keywords: Abductive Logic Programming, Agent Languages, CLIMA 2010, Intelligent Agents, Model-theoretic Semantics