王晓光,李梦琳,宋宁远.科学论文功能单元本体设计与标引应用实验[J].中国图书馆学报,2018,44(4):73~88
Design and Application of Scientific Paper Functional Units Ontology
科学论文功能单元本体设计与标引应用实验
Received:January 19, 2018  Revised:March 12, 2018
DOI:
Key words:Scientific papers  Functional units  Content ontology  Ontology construction  Deep indexing
中文关键词:  科学论文  功能单元  内容本体  本体构建  深度标引
基金项目:本文系教育部人文社会科学重点研究基地重大项目“大数据资源语义表示与组织”(编号:16JJD870002)和国家社会科学基金重大项目“基于认知计算的学术论文评价理论与方法研究”(编号:17ZDA292)的研究成果之一
Author NameAffiliationE-mail
WANG Xiaoguang 武汉大学信息管理学院 湖北 武汉 430072 wxguang@whu.edu.cn 
LI Menglin 武汉大学信息管理学院 湖北 武汉 430072  
SONG Ningyuan 武汉大学信息管理学院 湖北 武汉 430072  
Hits: 2417
Download times: 1184
Abstract:

    With the increasing of knowledge resources and the demand of knowledge mining, it is important to enrich the semantics of academic literature, which can not only help users quickly and accurately locate the knowledge units in scientific papers, but also can help readers to conduct comparative analysis and strategic reading. Therefore, it's essential to identify and describe the components and their semantic functions within scientific papers for promoting knowledge discovery and knowledge services.

Scientific paper content ontology is the standardized knowledge representation of scientific papers content structure and semantic function. It is of great significance for the deep indexing, information extraction and knowledge mining of scientific papers. After a review on the existing researches of paper components and attributions as well as the published ontologies, the existing ontologies, limited to the fundamental theories, have some deficiencies in revealing the deep semantics of information embedded in scientific papers. In order to design and build a component ontology, which is more suitable for information extraction, the functional unit theory should be considered.
The functional unit theory is the fundamental theory that combines information tasks and genre analysis, which is more suitable for the development of scientific paper content ontology oriented to knowledge discovery. Based on the functional unit theory, a novel ontology named Scientific Paper Functional Units Ontology(FUO)is designed. After reviewing the 41 functional units, 28 components are redesigned, including background, goal, motivation, method description, conclusion, contributions, etc. Based on the components, 12 classes and 28 subclasses are designed. The attributions of the classes are also designed by refering to Bio Event ontology and News Event ontology. The classes and attributions of FUO are formally represented with protégé 5.1. Then 10 research papers from JASIST are randomly selected to conduct a deep indexing experiment by using the GATE, a semantic annotation software. Finally, the distribution of different functional units within scientific papers is analyzed.

The originality of this research lies in the clear definition of the functional units with their attributes and the FUO which can reveal semantic features of scientific papers components in a more comprehensive and detailed manner. The results have also proved the potential availability of FUO for deep semantic indexing, semantic retrieval and knowledge discovery. This research deepens our understanding on scientific paper as a knowledge container from the perspective of information science. The limitation of this paper is the lack of considering the semantic relationships between content components of scientific paper. More detailed definition of the relationships and new components such as interactive tables, datasets, audios and videos should be studied in the future. 4 figs. 9 tabs. 47 refs.

中文摘要:
      科学论文内容本体是科学论文内容结构和语义功能的形式化和规范化知识表示,对于科学论文的深度标引和知识挖掘具有重要意义。本文系统梳理了已有科学论文内容表示模型和内容本体,并以功能单元理论为基础,提出科学论文功能单元本体的设计思路,构建包含28个类和5种属性在内的科学论文功能单元本体FUO。借助本体构建工具Protégé,对科学论文功能单元本体FUO进行形式化表示。借助语义标注工具GATE,利用功能单元本体FUO对论文进行初步的深度标引实验,检验了该本体的可用性。结果表明,功能单元本体FUO能够很好地表示科学论文内容组件的语义功能及其属性,揭示科学论文正文各部分的语义特征,可以用于面向知识发现的科学论文深度语义标引,为科学论文内容本体开发奠定了基础。图4。表9。参考文献47。
View Full Text   View/Add Comment  Download reader