一种改进的TFIDF网页关键词提取方法  

AN IMPROVED TFIDF-BASED APPROACH TO EXTRACT KEY WORDS FROM WEB PAGES

作　　者：李静月[1] 李培峰[1] 朱巧明[1] 

机构地区：[1]苏州大学计算机科学与技术学院,江苏苏州215006

出　　处：《计算机应用与软件》2011年第5期25-27,共3页Computer Applications and Software

基　　金：国家自然科学基金(90920004;60970056;60873150);江苏省自然科学基金(BK2008160);江苏省高校自然科学重大基础研究项目(08KJA520002)

摘　　要：传统TFIDF关键词提取方法虽然实现起来简单,时间复杂度低,但是效果并不理想,难以获得对文本内容起到关键性作用的特征。提出了一种在考虑中文文本结构特征和中文词语词性特征的基础上,借助扩展的同义词词林,利用改进的TFIDF公式来提取的方法。实验结果表明:该方法明显优于传统方法,能够抽取到令人满意的结果。Although the classical TFIDF-based keywords extraction method is easy to implement and has low time cost,its results are not good enough to acquire features which play key roles in the text content.This paper proposes such a method,it uses the improved TFIDF formula to extract keywords by means of extended synonym dictionary and based on the considerations of the structure feature of Chinese texts and the lexical category feature of Chinese words and phrases.Experimental results show that our method outperforms the classical method evidently,and is able to extract satisfied results.

关 键 词：文本结构 关键词抽取 TFIDF 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]