详细信息
文献类型:期刊文献
中文题名:基于术语语义相关的知识关联方法研究
英文题名:Knowledge Association Method Based on Semantic Relatedness
作者:李楠[1];孙济庆[1];吉久明[1];陈荣[1]
机构:[1]华东理工大学科技信息研究所
年份:2015
卷号:34
期号:6
起止页码:608
中文期刊名:情报学报
外文期刊名:Journal of the China Society for Scientific and Technical Information
收录:CSTPCD;;国家哲学社会科学学术期刊数据库;北大核心:【北大核心2014】;CSSCI:【CSSCI2014_2016】;
基金:国家社会科学基金项目(13BTQ053)
语种:中文
中文关键词:术语;词素;语义相关;结构化语义分析;知识关联
外文关键词:terminology, morpheme, semantic relatedness, structurized semantic analysis, knowledge association
摘要:针对知识发现服务系统对知识关联系统化、自动化的技术需求,提出一种基于结构化语义分析的术语知识关联自动构建方法。借助语言学的词素理论从微观层面剖析术语内部的语义结构规律,通过研究词素在术语语义表达中的功能和作用,探讨词素层面的术语结构化语义分析方法。基于该方法实现术语的知识度量及语义相似度计算,并设计术语知识关联的动态构建流程。最后,设计并实现了面向化学领域的知识关联实例,实验结果表明,本文提出的方法可以有效地建立符合领域知识背景的知识关联网络。
Knowledge association is an important research topic in knowledge discovery system, which needs to be solved systematically and automatically. This paper proposes an automatic knowledge association method of domain-specific term based on structurized semantic analysis. This method applies linguistic morpheme theory to analyze the semantic structure from the inner side of terminology on the micro level, focusing on the role and function of morpheme in knowledge expression. Then according to the above idea, improved knowledge value and semantic similarity computation algorithms are put forward and a dynamic process of term semantic association is realized. Finally, the knowledge associating experiment is conducted in chemistry, and results show that a scientific and reasonable term knowledge network can be built effectively.
参考文献:
正在载入数据...
