Abstract: Enterprise information retrieval has become the emphases and difficulty in the field of information retrieval，this paper discusses the trend and development of the core technology for enterprise information retrieval.An enterprise information retrieval system based on notion retrieval is designed and implemented，providing a query expansion algorithm and a self-learning algorithm to search the synonymous from special or user defined dictionary and search the co-appearance items from learning clusters.The system is adaptable and platform-independent.The independent architecture design contribute to the easily replacement of dictionary，which make the system can serve for enterprise in different industry with different platform，especially fit for the small and medium enterprise application.

Key words: query expansion, corporation information retrieval, information retrieval

摘要： 高效企业信息检索已成为信息检索领域的重点和难点，讨论了企业信息检索相关技术的发展，设计并实现了一个基于概念的企业信息检索系统，利用查询扩展算法对用户输入的关键词进行语义扩展：利用专业词典查找同义词，通过学习指定文档集合找出关联词，并允许用户自定义关联词进行扩展，从而实现真正意义上的概念搜索。系统设计充分考虑可适应性及平台无关性问题，其层次间独立的结构设计使得系统字典可替换，可用于不同行业不同平台的企业信息查询，特别适合中小型企业的轻型简便应用。

关键词: 查询扩展, 企业信息检索, 信息检索