【作者】 李俊；

【导师】 吴江；

【作者基本信息】 西北大学 ， 计算机应用技术， 2007， 硕士

【摘要】 目前，知识管理已经成为企事业单位提高技术、加快发展的重要手段。案例知识管理属于知识管理系统的一个方面，也逐渐成为研究的热点。而在传统案例知识管理模式下，由于信息缺乏统一的、通用的语义描述，用户难以找到所需的资源，而且用户难以向系统中加入在实践中获得的新的案例及知识。本文在分析现有案例知识管理系统的不足，以提高系统的动态参与性和通用性为目标，对本体论、元数据及基于案例推理等关键技术进行了研究，论文的贡献体现在下述几个方面：1)本文提出了一个基于本体的案例知识管理框架，该框架采用三层结构，即存储层、功能层和表现层，实现了案例知识管理的三个步骤，即知识的共享、知识的结构化及知识的传播。2)基于本体论和元数据标准建立了案例知识库，实现了对知识对象(概念、案例及资源)的表示、标注、存储、检索和更新。使用本体描述语言表示知识对象，较好的解决了语义上的同义及相似问题，从而实现了知识的共享、重用和互操作，提高了系统的动态参与性。3)对知识对象建立相应的知识项，屏蔽了知识对象内容的细节和表示形式的异构性，将知识对象的实体与用于检索的知识项分离，提高了检索速度。知识项还将案例与相关概念以及资源连接起来，增强了案例的理论基础和实用性。4)建立了较通用的案例模型，在案例的建模过程中，将案例的特征属性集封装起来，扩展了推理案例的适用领域。5)实现了对知识对象的多种检索方式，主要包括导航检索、语义检索、基于案例推理(CBR)的检索、全文检索和专家检索。6)本文实现了一个基于本体的案例知识管理系统原型，该系统采用B/S三层结构。更多还原

【Abstract】 In recent years knowledge management has become an important means of improving the technology and accelerating the development in many enterprises and institutions.Case knowledge management is a facet of the knowledge management system, which has become the research point. However, in the case of traditional knowledge management mode,as the lack of uniformity and universality, so users of knowledge management system can hardly find the necessary resources and hardly add the new case or knowledge which are acquired from practice.Based on the analysis of the disadvantages in the existing case knowledge management system,this thesis aims at the goal of improving the dynamic property and universality of case knowledge system.So that this thesis researches on the key technologies of case knowledge management system, such as ontology, metadata and case-based reasoning.The contributions of this thesis will be shown as follow:1) This thesis proposes a framework of ontology-based case knowledge management. The framework has three layers which are store layer, function layer and representation layer.The framework implements the three steps of knowledge management, which are the sharing of knowledge, the structurization of knowledge and the spread of knowledge.2) On the basis of ontology and metadata theory, this thesis builds the case and knowledge base and implements the representation, annotation, store, retrieval and update of knowledge object which contains concepts, cases and resources.The ontology is applied to represent the knowledge object, which solves the problems about semantic synonymy and similitude. The sharing, reuse and interoperability among the knowledge can be implemented by using of the ontology technology.Users can, meanwhile, participate in the system to update the knowledge base dynamically.3) This thesis builds the knowledge item for the knowledge objects, which hides the detail and heterogeneous of knowledge object and takes apart the knowledge item for retrieval from the entities of objects, so that speeds the knowledge retrieval.The knowledge item links the case, concept and resource, so the theory infrastructure and practicability of the cases can be advanced in this way.4) This thesis builds the universal mode of a great deal of cases, the feature attributes of case are encapsulated in moding process of case, which extends the practical domain of case.5) This thesis finishes the multiplex retrieval modes, such as navigation retrieval, semantic retrieval, Case-based Reasoning retrieval, full-text retrieval and expert retrieval.6) This thesis implements a system prototype which is Ontology-based Case Knowledge management. The system has adopted three layers of structure mode of B/S.更多还原