基于实体-关系-问题的Agent结构与推理模型  

Model for Structure and Reasoning of Agent Based on Entity-Relation-Problem

作　　者：张鹏[1] 王延章[1] 陈雪龙[1] 丁锋[2] 

机构地区：[1]大连理工大学管理与经济学部,大连116024 [2]大连理工大学软件学院,大连116620

出　　处：《情报学报》2013年第7期724-732,共9页Journal of the China Society for Scientific and Technical Information

基　　金：国家自然科学基金重大研究计划重点支持项目(91024029);国家自然科学基金青年项目(71203019);教育部博士点新教师基金项目(20110041120028);教育部人文社科基金项目(11YJC630023)

摘　　要：提出了一种新的知识表示方法，即实体一关系一问题（Entity—Relation—Problem，E-R-P）知识表示体系，进而给出了一种基于E—R—P知识表示体系的Agent结构与推理模型ERPBAM。该模型具有简单、直观、灵活、通用等特点，解决了以往Agent结构与推理模型因基于复杂的符号表示与推理而显得过于复杂的问题，并具有处理推理过程中涉及到的各类信息，尤其是模糊信息的能力。同时，由于E—R—P知识表示体系综合了客观事物系统及现实问题知识的表示，使得ERPBAM中Agent的结构与推理过程更好地融为一体，其实现代码变得高效而紧凑。构造了一种基于ERPBAM的智能Agent——ERPBIA。鉴于ERPBAM的特点，ERPBIA的构造过程变得简单，且具有以简单、高效的方式代替人类进行建模、推理、制定决策的能力。Firstly, a new knowledge representation system, named entity-relation-problem （E-R-P） knowledge representation system, is proposed. Then a model for structure and reasoning of agent based on the E-R-P knowledge representation system, named ERPBAM, is put forward. ERPBAM is straightforward and general. So it solves the problem of complexity of structure and reasoning for Agent, which is caused by complex symbol representation and deduction. Furthermore, ERPBAM has the ability to handle all kinds of information, especially the fuzzy information, involved in the reasoning process. Because E-R-P knowledge representation system synthetically represents the knowledge of objective system and actual problems, the structure and reasoning process of Agent in ERPBAM become more integrated, and the corresponding implementation code becomes more compact. Additionally, a new type of agent, called ERPBIA, is constructed based on the ERPBAM. In view of the features of ERPBAM, ERPBIA can be constructed easily, and it also can conduct modeling, reasoning and making decision on behalf of human beings with simple, efficient style.

关 键 词：智能AGENT 知识表示 实体-关系-问题 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]