基于案例推理的知识系统的设计与实现

【作者】 刘健；

【导师】 陈前；

【作者基本信息】 南京航空航天大学 ， 计算机软件与理论， 2004， 硕士

【摘要】 近来，随着计算机网络和并行技术的发展，人工智能再次成为人们研究的热点，基于案例的推理方法也越来越受到人们的重视。与传统的基于规则的系统相比，基于案例推理的系统具有简化知识获取、便于知识积累等优点，在没有很强的理论模型、领域知识不完全、经验知识丰富的故障诊断、决策支持等领域得到了广泛的应用。 本文介绍了基于案例推理方法的一般知识。讨论了基于案例推理的知识系统的设计和实现，并具体讲述了作者所实现的一个基于案例推理的知识系统。 本文针对复杂系统具有庞大案例库的特点，提出了一种通过建立多级案例库语言编译器的方法来建立多级案例库，并以此来分解庞大案例库的方法。 本文实现了一个多级案例库语言编译器，并在此基础上实现了一个支持多级案例库的基于案例推理的知识系统。更多还原

【Abstract】 Recently, with the development of computer network and parallel technology, Artificial Intelligence has become the hot research topic again and Case-based Reasoning technology has received more and more attention. Comparing to traditional Rule-based System, Case-based Reasoning System has many advantages, which can predigest knowledge gathering and easily accumulate knowledge. Case-based Reasoning technology has been widely used in fault diagnosis and decision-making support which abound of experience knowledge but has neither good theoretic model nor complete field knowledge.In this paper, the author introduced the basic knowledge of Case-based Reasoning technology and the general knowledge of how to construct a knowledge system using Case-based Reasoning technology. Then, the implementation of a Case-based Reasoning system is carried out in details.The casebases of modern large-scale engineering systems are likely too large to maintain and utilize efficiently. In this paper, a hierarchical casebase is proposed to tackle this problem; the hierarchical casebase structure should streamline the searching and reasoning process.In this thesis, the author has implemented a Hierarchical casebase compile system and its case-matching algorithm for this system.更多还原