摘要
结合搜索引擎搜索结果集的特点,应用中文分词技术、关联分析来分析词与词之间的关联关系以及名词短语词性过滤,最终从特定主题的搜索结果集中挖掘出该主题的相关概念,提出了一种自动挖掘相关概念的方法,实现了一个T op icM iner(简称TM)原型系统,并通过实验证明了该方法的可行性,概念挖掘的准确率达到了60%以上。
This paper puts forward a novel and automatic method of mining relative concept, which uses Chinese textsegmentation technique, association analysis , semantics knowledge to get the relative concept. Topic Miner, a prototype system, is implemented and proves the effectiveness of the method mentioned above. The experiment shows the accuracy of concept mining is higher than 60%.
出处
《电脑开发与应用》
2007年第2期14-16,共3页
Computer Development & Applications
关键词
概念挖掘
关联分析
频繁项集
概念检索
concept mining, association analysis, frequent items , concept retrieval
