利用关联规则挖掘文本主题词的方法  

Approach for Extracting Thematic Terms Based on Association Rules

作　　者：刘菲[1] 黄萱菁[1] 吴立德[1] 

机构地区：[1]复旦大学计算机科学与工程系,上海200433

出　　处：《计算机工程》2008年第7期81-83,共3页Computer Engineering

摘　　要：主题词抽取是目前信息检索领域研究的热点,与一系列数据挖掘相关的任务密切相关。该文提出一种新的利用关联规则挖掘中文文本主题词的方法,该方法抽取的主题词包括关键词和相关检索词两部分。在关键词抽取的基础上,采用数据挖掘中的关联规则挖掘算法抽取相关检索词,用于扩展检索或相关检索,提高了用户对于文档的理解。实验表明该方法取得了较好的效果。Thematic terms extraction is one of the hot topics in the field of information retrieval, and in tight relationship with a variety of data mining tasks. This paper presents an approach for extracting thematic terms based on association rules, which include both keyphrases and related terms. Based on keyphrase extraction, related terms could be used in extended search and related information retrieval, as well as providing users with a better understanding of the topic. Experimental results show that the method is effective in extracting both keyphrases and related terms.

关 键 词：关键词抽取 关联规则挖掘 文本挖掘 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]