摘要
基于大规模层级分类语料库,抽取网页上专家已经标引的关键词形成关键词表;针对关键词的领域不均匀性和邻界域两个特征,提出并模拟计算了关键词表征文本主题特征程度的主题度。以关键词及其主题度为领域知识,结合统计方法,完成了一个知识与统计相结合的关键词自动标引系统。
From a classed large-scale corpus, extracts keywords labeled on web pages by indexing specialist and formed a keywords list; Referring to the two characteristics of keywords: fields non-even and exists range edge, brought up and calculated the words' subject degree by statistical model. Subject degree expresses text content' s subject concept. Based on subject degree, constructed a key words auto-indexing system.
出处
《图书情报工作》
CSSCI
北大核心
2007年第7期41-43,共3页
Library and Information Service
关键词
关键词标引
分类语料库
主题度
keywords indexing classed corpus subject degree