Abstract: CDSE, a model for domain-based intelligent search engine is proposed. The model can help people to retrieval what they need by combining text classification with key phrase extraction. Several algorithms that use key technology are proposed, such as statistics, data mining and agent. These algorithms solve the shortcomings effectively of low precision and relevant document ranking behind in now search engine.

Key words: Intelligent search engine, Key phrase extraction, Data mining, Agent

摘要： 介绍了一个面向领域的智能搜索引擎CDSE(Computer Document Search Engine)的设计和实现。CDSE结合文本分类和关键词组抽取检索用户需要的信息。利用了多个算法，综合运用了统计学方法、数据挖掘技术和Agent技术，较好地解决了现有搜索引擎普遍存在的搜索精度差、相关文档列序较后的问题。

关键词: 智能搜索引擎, 关键词组抽取, 数据挖掘, Agent