基于HNC理论的期刊文献知识组织研究

发布时间：2018-02-22 04:21

本文关键词： 知识组织 HNC(概念层次网络)理论知识元词汇链　出处：《大连理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：网络通讯技术的飞速发展使得人们可以接触的信息呈爆炸式增长,急需的知识通常被实际不需要的信息垃圾淹没,信息时代竞争力的关键不是实物资产或资源,而是知识,所以要像其他资产一样管理和利用知识。针对期刊文献的组织管理,有必要深入文献内部,对文献知识合理表示、整理和关联,使得各学科各领域形成相互联系的语义网络。本文首先分析了本课题的研究背景和国内外研究现状,明确本文的研究对象是期刊文献,目标是期刊文献知识的有效管理,以便为用户提供更好的知识服务。然后阐述了知识组织与管理的相关理论和技术,在此基础上针对期刊文献的知识结构分析,分别从文献知识的表示、抽取、整理和关联几个角度对其组织管理。第一,对期刊文献结构分析,把逻辑结构划分为主题、子主题、知识元、句子；提出期刊文献的知识元六元组的描述模型,把词汇链引入作为知识元属性,分析每一属性的涵义和对知识关联的作用；提出基于HNC理论的词汇链构建方式,以便为下文中知识元的组织服务。第二,以词汇链为基础,抽取知识元的主题词和内容,在词汇链权重基础上,计算知识元主题词和内容的权重影响因子,从准确性和覆盖率两方面确定抽取比例；对知识元的主题领域划分提出利用HNC的领域概念归类,使同一领域粒度下的知识元汇集在一起。最后基于知识元的属性和HNC理论,分析知识元的逻辑关联和语义关联,描述了由词汇链、主题领域、HNC概念关联式对知识元之间的语义链接作用。第三,针对不同领域粒度下的知识元建立相应的知识元库,以情报知识元为例检索,说明知识元检索的意义和价值。本文主要是对文献内部知识的合理表示和组织,为用户呈现深入文献内部的知识,迅速了解知识发展状态和学科分布,促进知识融合、创新,为个性化知识服务提供新思路。
[Abstract]:With the rapid development of network communication technology, the information that people can contact increases explosively. The badly needed knowledge is usually submerged by the information garbage that is not needed. The key to competitiveness in the information age is not physical assets or resources, but knowledge. Therefore, to manage and utilize knowledge as other assets, it is necessary for the organization and management of periodical documents to go deep into the literature, to reasonably express, organize and correlate the literature knowledge. This paper first analyzes the research background of this subject and the current research situation at home and abroad, and clarifies that the research object of this paper is periodical literature, and the goal is the effective management of periodical literature knowledge. In order to provide users with better knowledge service. Then the related theory and technology of knowledge organization and management are expounded. Based on this, the knowledge structure of periodical literature is analyzed and extracted from the representation and extraction of literature knowledge. Firstly, the logical structure is divided into themes, sub-themes, knowledge elements, sentences, and a description model of six tuples of knowledge tuples of periodical literature is proposed. This paper introduces the lexical chain as the attribute of knowledge element, analyzes the meaning of each attribute and its effect on knowledge association, and puts forward the construction method of lexical chain based on HNC theory in order to serve the organization of knowledge element in the following. Second, it is based on the lexical chain. On the basis of the weight of lexical chain, the factors affecting the weight of the subject words and contents of knowledge elements are calculated, and the extraction ratio is determined from the two aspects of accuracy and coverage. Based on the attribute of knowledge element and HNC theory, this paper proposes to classify the domain concept of HNC to make the knowledge element gather together. Finally, based on the attribute of knowledge element and HNC theory, the logical association and semantic association of knowledge element are analyzed. This paper describes the semantic link between knowledge elements by lexical chain, topic domain HNC concept association. Thirdly, the corresponding knowledge element database is established for different domain granularity, and the information knowledge element is retrieved as an example. This paper is mainly about the rational representation and organization of the internal knowledge of the literature, presents the knowledge inside the document for the user, quickly understands the knowledge development state and the discipline distribution, promotes the knowledge fusion, the innovation, and so on. To provide new ideas for personalized knowledge service.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：G254;TP391.1

【参考文献】