【作者】 李衍淼；

【导师】 梁永全；

【作者基本信息】 山东科技大学 ， 计算机软件与理论， 2006， 硕士

【摘要】 随着信息化的不断发展，人类的信息生产能力高速增长。如何在浩如烟海的信息中找到对于用户有价值的信息已经成为一个非常重要的问题。不断发展的信息检索技术使这一问题的解决成为了可能。 近来，基于本体的信息检索技术成为了研究的热点。本体论从本质上讲是就是对客观存在的概念和概念之间关系的描述，所以基于本体的信息检索是基于知识的、语义上的检索。它弥补了传统的基于关键字信息检索技术的缺陷，从而在查准率和查全率上有更好的保证。 本文首先对本体论技术进行了介绍，为基于本体的智能检索系统的设计实现提供了理论基础。本体是共享的概念模型的形式化的规范说明，能够以一种明确的、形式化的方式来表示领域知识，促进知识共享。本文对各种本体描述语言进行了比较与分析，并对现有的本体构建方法论进行了介绍。 其次，对应用到智能检索系统的本体论中的关键技术进行了研究。本文对现有领域本体建设方法进行了分析与比较，将软件工程中螺旋原型法的思想应用到领域本体建设之中。然后针对智能查询问题对基于本体的概念相似度和相关度的算法进行了研究，提出了概念相似度和相关度的一种量化方法。 最后，本文通过对一个现有信息检索系统的分析提出了一个基于本体的多主体信息检索模型。通过对本体技术的应用，系统实现了自动分析文档的领域属性和智能化地规范和显化用户的查询要求的功能，从而改善了系统的信息检索的效率，并提高了检索的查准率与查全率。更多还原

【Abstract】 With the development of information society, the information productivity growths at high speed. How to find valuable information for people in the tremendous amount of information has become an extremely important problem. The developing information retrieval technology provides a solution to the problem.The information retrieval technology based on ontology is the emphasis of the present research. In essentially, ontology is the description of the relations between objective existing concepts, so the information retrieval technology based on ontology provides semantic matching which makes up for the flaw of the traditional information retrieval technology based on keywords. It is better at recall and precision of the query result.First, the thesis introduces the ontology technology. It provides a theoretic foundation for the implementation of the information retrieval system based on ontology. Ontology is a formal specification of a shared conceptualization and it can express domain knowledge in a specific and formal way, accelerating information sharing. This paper compares and analyzes ontology describing languages, and it also introduces present ontology constructing methodology.Secondly, the paper studies the key ontology technology used in intelligent information retrieval system. This thesis compares and analyzes the present ontology constructing methodology, then applies the ideology of spiral rapid prototype method to the domain ontology constructing. It studies the algorithm of similarity and relevance of conceptions for the problem of intelligent query. This solves the problem of calculation with traditional knowledge.At last, in this paper, an information retrieval system based on ontology and multi-agent is presented by analyzing an existing更多还原