文章摘要

鲍秀林,吴雯娜.语义映射质量及影响因素分析——以国家叙词库语义映射为例[J].中国图书馆学报,2016,42(5):57~67
语义映射质量及影响因素分析——以国家叙词库语义映射为例
Analysis of the Quality of Semantic Mapping and Its Influencing Factors:A Case of the National Thesauri Warehouse's Semantic Mapping
投稿时间:2016-03-17  
DOI:10.13530/j.cnki.jlis.160023
中文关键词: 语义映射  多表映射  等同映射  映射质量  国家叙词库
英文关键词: Semantic mapping  Multi thesaurus mapping  Equivalent mapping  Mapping quality  The National Thesauri Warehouse
基金项目:本文系国家社会科学基金项目“国家叙词库构建方式与发展机制研究”(编号:13BTQ013)的研究成果之一
作者单位E-mail
鲍秀林 中国科学技术信息研究所 北京100038 baoxl926@163.com 
吴雯娜 中国科学技术信息研究所 北京100038  
摘要点击次数: 2927
全文下载次数: 1052
中文摘要:
      国家叙词库采用以《汉语主题词表》为中心表、其他中文叙词表向中心表映射的模式进行基于概念的语义集成。为了研究学科交叉对国家叙词库语义映射造成的影响,本文选取八部叙词表中环境科学大类5 120个概念,与《汉语主题词表》环境科学大类8 329个概念进行映射实验。通过计算机自动匹配发现,多表映射存在一个概念等同映射到多个概念,多个概念等同映射到一个概念等逻辑问题。本文以实例形式分析这些逻辑问题的存在形式,并从概念专指度、概念语义模糊、语义关系结构、学科背景等方面分析产生这些逻辑问题的主要原因;在此基础上,提出解决这些逻辑问题的原则以及映射完整性的判定规则。图16。表5。参考文献14。
英文摘要:
The National Thesauri Warehouse is an integrative service system of the thesaurus resources of China. Its construction is a large and complex project,which includes many aspects while its core part lies in the semantic integration of multi thesaurus. Based on the concept of semantic integration,the National Thesauri Warehouse takes Chinese Thesaurus as the center vocabulary and uses other Chinese thesaurus mapping to the center vocabulary.
To explore the effect of interdisciplinary on the National Thesauri Warehouse semantic mapping,we select 5 120 concepts from the environmental science classes of 8 thesaurus and 8 329 concepts from the environmental science class of Chinese Thesaurus to take mapping experiments.
By analyzing the results of automatic matching,we find that the existence of quasi synonyms,related words,hyponyms/hypernyms in words of equal relationships,especially inconsistent mapping methods of two thesaurus for synonyms,causing many logical problems in multi thesauri mapping,such as one concept equivalent mapping to multiple concepts or multiple concepts equivalent mapping to one concept. Since the notion of equivalent mapping is the foundation for recognizing and processing other mapping relationships,it can be further developed to hierarchical mapping and relevant mapping,and thus to enrich the network relationships among the conceptual nodes. Therefore,the main purpose of this paper is to study the logical problems generated by equivalent mapping.
In this paper,we analyze the forms of the logical problems and the causes at different dimensions,such as concept specificity,conceptual semantic fuzziness,semantic relationship structure,subject background. In view of these,the principles of solving these problems are put forward,and the decision rules of mapping integrity are also presented.
Semantic mapping of different thesaurus has many advantages including not only of great help to the compatibility and semantic specification,and improve the quality of the National Thesauri Warehouse,but also enriching and supplementing the semantic relation of Chinese Thesaurus,and promoting the integrated application of Chinese thesaurus. The analysis of the logical problems of semantic mapping and its influencing factors may have a reference value to the actual mapping work can be provided. The weakness of this paper lies in the limited experimental data,where there may be some logical problems that have not been completely found. 16 figs. 5 tabs. 14 refs.
查看全文   查看/发表评论  下载PDF阅读器