本体概念匹配技术的研究与实现-计算机软件与理论专业论文.docxVIP

摘 要 本体作为语义网中的知识表现形式，近年来已经被广泛的应用到知识工程、人 工智能和信息检索等研究领域。由于不同的组织或个人在本体构建中没有统一的 标准，导致了本体异构的问题，所以需要对本体进行映射、合并和集成。本文的 本体概念匹配技术主要服务于本体映射、合并乃至对本体的集成。 本体概念匹配是将不同本体中的概念进行匹配，以实现本体映射、合并和集成， 消除本体异构、实现语义通信和互操作等过程。本文首先研究本体概念匹配的关 键技术——中文语义相关度计算方法，分析了英文维基百科 ESA 算法，将其移植 到中文维基百科上，并对其进行了重定向词合并和链接词加权两方面的改进，通 过实验证明改进后的算法提高了中文语义相关度计算的准确性和效率。然后研究 本体概念匹配方法，分析了二分图匹配的稳定婚姻问题和延迟认可算法，将其移 植到本体概念匹配上，并对延迟认可算法进行了改进。最后基于上述两个关键技 术，设计并实现了本体概念匹配工具来验证本文研究的可行性。 实验表明，本体概念匹配工具的匹配结果稳定、合理，符合本体映射、合并和 集成的要求，达到了预定的研究目标。 关键词：本体 概念匹配 显式语义分析 维基百科 Abstract As the expression modle semantic net, ontology has been widely applied in knowledge engineering, artificial intelligence, information retrieval and other research fields. Since different organizations or individuals obey different rules in ontology architecture, there are different ontologies in different domains. Therefore, the mapping, mergence and integration of ontologies are required. The Ontology Concepts Matching (OCM) technique presented in this dissertation will mainly make effect in ontology mapping, ontology mergence and even ontology integration. Via matching the concepts in different ontologies, OCM realizes the mapping, mergence and integration of ontologies, eliminates the differences of ontologies in different domains, and implements semantics telecommunication, mutual operation and other processes. In this dissertation, first, the core technology of OCM Chinese semantics correlativity computation algorithm is studied; second, the ESA algorithm of Wikipedia is analyzed and expanded in Chinese Wikipedia, and the improvements are made in redirect word mergence and weighting of connected words, moreover, by applying the improved algorithm the Chinese semantics correlativity computation accuracy and efficiency is proved to be enhanced via experiments; finally, based on the two key techniques presented above, a OCM testing tool is designed and implemented to test the feasibility of the research in this paper. Experiments manifest that the result of OCM