面向知识服务的知识库结构研究

High efficiency knowledge discovery, as supported by the process from information service to knowledge service which to solve actual problems, put forward to effectively promote knowledge service and widen the service field. The aim of our project is to form the model of knowledge base structures on architecture from the perspective of knowledge service, to construct the theoretical framework of knowledge base structures driven by the knowledge discovery, and to protect high performance knowledge service. The research of this project does focus on the logical structure and frame structure of knowledge base, exploring the general law of knowledge base structures, and the formation of a new pattern of knowledge discovery. About the logical structure of knowledge base research is according to the following aspects: investigate the mapping relationship between knowledge points and raise a new simple knowledge representation based on reasoning ranges, settle a dispute between knowledge storages and knowledge reasoning, safeguard the utility of the simple knowledge representation by introducing the mechanism of data cleaning, obtain query results with a clean degree satisfying clean requirements of applications from knowledge base with the dirty data. The project also research on how to analysis the frame structure of knowledge base. By breaking the tradition closed structure of knowledge discovery based on database, a double-base cooperating is proposed for gaining deeper knowledge in the whole process of the knowledge discovery. The heuristic coordinator algorithm and interruptive coordinator algorithm are given to guarantee double-base cooperating normal operation and to evolve the frame structure of knowledge base by itself.

由传统的文献信息服务，提升到用于解决实际问题的知识服务，必须以高效的知识获取作为支撑，有效地推动知识服务并拓宽其服务领域。本课题拟从知识服务的角度架构知识库结构模型，并以知识获取为驱动来构建知识库结构的理论框架，保障知识服务水平的提升。本课题重点研究知识库的内在逻辑结构与整体框架结构，探索知识库结构一般规律，形成知识获取的新模式。分析知识库内在逻辑结构，根据推理范畴探究知识点之间存在的映射关系并提出知识的简约表示，解决了知识存储与知识推理两难问题；引入数据非清洁度的机制，保障了知识的简约表示的效用，使得能从知识库中获取满足非清洁度的查询结果且适应不同知识服务的合理化需求。探究知识库整体框架结构，突破基于数据库知识获取的封闭结构，提出将知识库、数据库双库有机融合并协同工作在知识获取的全过程，获得更深层次的知识，并给出了启发协调算法与中断协调算法，使知识库整体框架结构结构具备了自我进化的能力。

本项研究从知识服务的需求出发，以知识服务的形式和要求为依据，从宏观上研究知识库的整体框架结构，从微观上探讨知识库的内在逻辑结构，并探讨了知识组织的理论基础与理论支撑，从应用角度阐释知识库中各类知识组织在技术上的实现和应用。并将理论研究的成果与结论用于应急决策领域的知识库结构的构建，将知识库结构体系研究凝练的一般规律在应急决策特定领域获得了实践应用和推广。系列的研究工作集中在设计了主题词表和分类表及其组织结构，探讨了语义信息的知识组织、复杂网络的深层组织、知识地图的数据组织，探究了知识点之间的映射关联、知识的简约表示，解决了知识存储与知识推理两难问题，研究了数据非清洁度的机制，并构建了知识库中满足非清洁度表示的知识服务模型，突破了基于数据库知识获取的封闭结构，提出了将知识库、数据库双库有机融合、协同工作的知识获取模型，研究并给出了启发协调算法与中断协调算法。研究工作不仅传承经典的知识组织的理论方法，更从知识库的研究视角创造和完善知识组织技术，使之适合于知识服务的需要，特别面对突发事件的应急决策领域，所构建的知识库结构体系更在特定领域中的知识服务中发挥作用，并使其有序地收敛于高效的知识服务。