面向非相关文献的知识关联发现系统设计与实现  

Design and Implementation of Knowledge Linkage Retrieval System for Non-Interactive Literature

作　　者：刘爱琴[1] 安婷 Liu Aiqin;An Ting

机构地区：[1]山西大学经济与管理学院

出　　处：《国家图书馆学刊》2019年第3期75-83,共9页Journal of The National Library of China

基　　金：山西大学人文社会科学科研基金项目“基于跨界思维的信息咨询新业态研究”(项目编号:115546003)的研究成果之一

摘　　要：面向非相关文献的知识关联能够促进新知识的产生,为科学研究寻找新的线索提供了一种便捷、有效的辅助手段。本文首先设计了面向非相关文献的知识关联发现系统,该系统以《中国分类主题词表》为主题词受控词表,通过对文献摘要的中文分词处理,提取主题词并标引,提取文档特征矩阵并利用计量分析技术和聚类技术分析文献间特征的相似、相异水平,从词的粒度层面对非相关文献之间的关系进行挖掘,揭示非相关文献的知识关联。其次,基于该系统为用户精确匹配相关的知识库,采用TOP-K算法反馈与用户相关的文献集,为用户提供满意度更高的知识发现及相关扩展服务。Knowledge linkage for non-interactive literature can promote the generation of new knowledge and provide a convenient and effective supplementary means for new clues of scientific research. This paper first designs knowledge linkage retrieval system for non-interactive literature, which uses the Chinese Classification Thesaurus as the subject word controlled vocabulary, with the Chinese word segmentation of the document abstract and the extraction of the subject words for indexing, extracting the document feature matrix and using the econometric analysis technique and clustering technique to analyze the similarity and difference level between documents, the relationship between non-interactive documents being mined from a more granular level, revealing knowledge linkage of non-interactive literature. Then, based on relevant knowledge base used to accurately match for users, this paper adopts TOP-K algorithm to feed back users-related document collections, so as to provide users with higher satisfaction knowledge retrieval and related extension services.

关 键 词：非相关文献 知识关联 中国分类主题词表 计量分析技术 知识发现 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] G354[自动化与计算机技术—计算机科学与技术]