论文部分内容阅读
1 引言现在不少人工智能(AI)研究者认为定义人工智能的一种方法就是将它看作是以构建具有智能行为的agent为目标的研究领域。从这种观点看,“agent”实际上就是人工智能的核心。自从80年代后期以来,关于agent理论及应用的研究取得了很大发展,目前agent是主流计算机科学,包括数据通讯、并行系统、机器人、用户接口设计等等的研究人员所讨论的一个主题。构建agent的经典方法是将它们看作特殊类型的基于知识的系统,这就是通常的符号方法,而相应的a-gent则被称为慎思的agent。一个慎思的agent包含对环境的确切描述的符号模型,并且其决策(例如应采取什么行动)都是基于模式识别或符号处理,通过逻辑推
1 INTRODUCTION Nowadays many artificial intelligence (AI) researchers think that one way to define artificial intelligence is to think of it as a research field aimed at constructing an agent with intelligent behavior. From this perspective, “agent” is actually the heart of artificial intelligence. Since the late 1980s, great progress has been made in the research on agent theory and its applications. Currently, the agent is a subject discussed by mainstream computer science researchers including data communications, parallel systems, robotics, user interface design, and the like. The classic way to build an agent is to think of them as a special type of knowledge-based system, which is the usual symbolic approach, and the corresponding a-gent is called a deliberate agent. A thoughtful agent contains a well-defined symbolic model of the environment, and its decisions (such as what actions should be taken) are based on pattern recognition or symbolic manipulation, pushed logically