基于词汇链的关键词自动标引方法  

Keywords Automatic Indexing Based on Lexical Chains

作　　者：李纲[1] 戴强斌[1] 

机构地区：[1]武汉大学信息资源研究中心,武汉430072

出　　处：《图书情报知识》2011年第3期67-71,共5页Documentation,Information & Knowledge

基　　金：国家自然科学基金项目(70673070)研究成果之一;"武汉大学2008年博士研究生(含1+4)自主科研项目"资助

摘　　要：关键词自动标引是一种识别有意义且具有代表性片段或词汇的自动化技术。关键词自动标引可以为自动摘要、自动分类、自动聚类、机器翻译等应用提供辅助作用。本文利用基于知网的词语语义相关度算法对词汇链的构建算法进行了改进,并结合词频和词的位置等统计信息,进行关键词的自动标引。实验证明,该方法可以有效的进行关键词的自动标引。keywords automatic indexing is an automation technology to identify meaningful and representative phrases or words.Keywords automatic indexing is broadly used in many applications such as automatic summarization,text classification,text clustering and machine translation.In this paper,we improved the lexical chain construction algorithm using semantic relevance computation algorithm of Chinese words based on HowNet.And combined with words statistical information such as words frequency and location of words,this algorithm is used in keywords automatic indexing.The method is fully considered all the various semantic relations between words and improved the performance of keywords automatic indexing.

关 键 词：语义相关 词汇链 关键词抽取 自动标引 

分 类 号：G356.6[文化科学—情报学]