节点文献

基于英汉双语语料库的英汉双语Ontology的建立与管理

Establishment and Management of the English-Chinese Ontology Based on the Bilingual Corpus

【作者】 黄新艳

【导师】 徐建良;

【作者基本信息】 中国海洋大学 , 计算机应用技术, 2006, 硕士

【摘要】 随着Internet的迅猛发展,信息技术面临的种种挑战促成了建立Ontology的热浪。通常,一个成熟本体的建立除了需要领域专家的参与外,还需要遵循Ontology的一系列构造准则和评估标准,成本十分高昂,其效果也不尽人意。目前,国外已建立了许多比较成熟的英语Ontology,国内也已有了比较完善的英汉双语语料库,现在完全可以利用这些已经成熟的资源来建立为我们自己所用的英汉Ontology。本文基于英汉双语语料库,实现了一种将英语Ontology转化为我们自己所用的英汉双语Ontology的方法,使我们能够以较低的代价获得比较成熟的英汉双语Ontology,这必将带动我国Ontology建设乃至其相关领域的飞跃发展。第一章介绍Ontology和双语语料库以及本课题的研究背景和研究目的。第二章介绍Ontology的一些重要概念,包括Ontology的定义、分类、应用、构建方法、开发支撑工具以及描述语言。第三章介绍语料库和双语语料库的概念及各种句子、短语对齐技术,首先回顾语料库的发展背景,然后比较双语语料库的各种句子、短语对齐技术的优缺点,在此基础上引入本人对英汉双语语料库中短语对齐的改进工作,并利用四种常见的数学模型建立了最优翻译等价对库。第四章给出基于英汉双语语料库的英汉双语Ontology的实现。利用最优翻译等价对库,采用Jena解析RDF描述的英文Ontology,用java语言实现英文Ontology转换为英汉Ontology。第五章从实现方法的创新性、局限性及可以进一步开展的工作等方面做一简单总结和展望

【Abstract】 Ontology has gained great progress with the fast development of the Internet and the chanlleges faced by Information Technology. Usually, Ontology is constructed through field experts and must obey series of Ontology rules. Each of these methods has its limitation. A lot have to be paid for these Ontologies and they are always unsatisfied .As a mature thing, Ontology technology abroad together with demestic English-Chinese bilingual corpus, will become the important resourses to construst the English- Chinese Ontologies for our own.In this paper, a method has been realized to transform English Ontology into English-Chinese Ontology, so that we can have an English-Chinese Ontology with high quality and dramatically low cost.Chapter 1 gives an introduction to Ontology and Bilingual Corpus, and provides a background and a motive for study.Chapter 2 gives some important concepts about Ontology, such as Ontology defination, Ontology taxonomy, Ontology applications, Ontology construct methods and Ontology description language.Chapter 3 gives some important concepts about bilingual corpus and methods about sentence parallel and phrase parallel, observes a historical overview, compares the merits and demerits of some mainly parallel methods. After that, this paper gives a bettered method about English-Chinese phrase parallel and proposed to construct a translation equivalences table through 4 co-occurrence based models.Chapter 4 shows an implementation of the transformation between English Ontology and English-Chinese Ontology based on bilingual corpus, where the translation equivalences table, java program language, and Jena serves as a RDF parse

【关键词】 本体双语语料库RDFJena
【Key words】 Ontologybilingual corpusRDFJena
  • 【分类号】TP392
  • 【被引频次】2
  • 【下载频次】363
节点文献中: 

本文链接的文献网络图示:

本文的引文网络