可重构知识管理系统中案例与规则的集成推理方法  

CBR AND RBR INTEGRATED METHOD IN RECONFIGURABLE KNOWLEDGE MANAGEMENT SYSTEM

作　　者：张兴[1] 王英林[1] 张申生[1] 

机构地区：[1]上海交通大学计算机科学与工程系,上海200240

出　　处：《计算机应用与软件》2008年第2期1-3,20,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(60374071);国家863/CIMS计划项目(2002AA411420);上海市科委科技攻关项目(03DZ19320)。

摘　　要：传统上的知识管理工具往往在设计时就固定了知识结构,这样的系统不但缺乏通用性,而且也限制了知识检索与推理的效率。介绍一个基于本体的可重构知识管理系统,知识作为本体概念的对象实例,利用本体模型的可定制性,解决了以往知识类型不能扩展的问题。详细阐述了结合案例推理与规则推理的集成推理方法,通过规则学习算法支持了规则库的动态扩充与调整,并将本体类的语义关系应用于推理方法,进一步提高了推理的效率。最后介绍了该系统在某飞机设计研究院的应用情况和今后的研究方向。Most traditional knowledge management tools are designed to be built on a fixed knowledge structure, so they lack generality and the efficiency of knowledge retrieval and reasoning is reduced, An ontology-based recontignrable knowledge management system is introduced. In this system, knowledge instances are represented as the objects of ontology concepts;and the previous problem of fixed types of knowledge is solved through the customization of ontology model. Then ,a CBR and RBR integrated reasoning method is discussed. Dynamic enlargement and adjustment of rule base is realized by rule learning algorithm. Moreover, semantic relations between ontological classes are applied to this method,which significantly improves the reasoning efficiency. Finally, the application of this system for an airplane design institute is introduced, and the direction of future research is pointed out.

关 键 词：知识管理系统 基于案例推理 基手规则推理 集成推理 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] F270.7[自动化与计算机技术—控制科学与工程]