基于词汇链的中文新闻网页关键词抽取方法  

Keyword Extraction Based on Lexical Chains for Chinese News Web Pages

作　　者：胡学钢[1] 李星华[1] 谢飞[1,3] 吴信东[1,2] 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009 [2]Department of Computer Science,University of Vermont,Burlington,VT 50405,USA [3]合肥师范学院计算机科学与技术系,合肥230061

出　　处：《模式识别与人工智能》2010年第1期45-51,共7页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金资助项目(No.60573174)

摘　　要：词汇链是一种词语间语义关系引起的连贯性的外在表现,提供关于文本结构和主题的重要线索.文中在解决歧义消解问题的基础上提出利用词汇链,结合词频特征、位置特征和集聚特征抽取中文新闻网页关键词的方法.该方法根据词语在文档中语义联系将文档表示成词汇链形式,并在此基础上抽取关键词.对中文新闻网页和学术期刊文献两种语料进行实验,结果表明该方法可明显提高抽取的关键词质量.A lexical chain is an external performance consistency by semantically related words of a text, and it is the representation of the semantic content of a text. Based on the word ambiguity resolution, a method for keyword extraction from Chinese news web pages is proposed by using lexical chains combined with frequency features, location features and cohesion features. The document is represented as lexical chains by the relationship between phrases and the key phrases are extracted from the lexical chains. The proposed method is tested on the corpus of Chinese news web pages and journal articles. The experimental results show that the proposed method improves the quality of the keywords extraction.

关 键 词：词汇链 关键词抽取 歧义消解 语义相似度 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]