摘要： 介绍了一种自动从因特网上抽取重要新闻的方法,以及按此方法研制而成的软件工具:NewsGrouper。它利用超文本中的信息,对超文本所指向的网页进行聚类分析,进而得到当日的重要新闻。该方法的优点是只需要用户提交信息源,不需要其他人工的干预,自动化能力高。因此,类似的技术也可以应用到文献的自动分类、检索等领域中。

关键词: 信息抽取, 文本聚类, 字符串距离

Abstract: In this paper, we present a new mechanism to extract important news from Internet, and a software tool named NewsGrouper, which is developed based on this mechanism. NewsGrouper can retrieve real-time important news by clustering and analyzing Web pages linked by HTML protocol, and by taking advantage of the information in the hypertexts. An advantage of this mechanism is its high automation, which means that the user only need to input information source without any further interaction. So, the same technology can also be used in other fields such as automatic document classification, retrieval and so on.

中图分类号: 

• TP311.5