【作者】 马静；

【导师】 詹海生；

【作者基本信息】 西安电子科技大学 ， 教育技术， 2010， 硕士

【摘要】 信息技术的迅速发展和万维网的出现产生了大量的数据,特别是以doc格式、pdf格式、txt格式的数据呈海量增长的趋势。如何合理的组织和存储这些异构数据,提高检索的效率,是全文检索领域面临的一个重要问题。本课题的研究任务是设计实现一个中文全文检索系统,实现对地方志资料的全文检索功能。本课题深入研究了全文检索的原理,包括索引定义、索引内容、索引创建和索引搜索。本课题采用Lucene开源搜索技术。本文详细分析了Lucene的结构组成、索引过程、搜索过程、数据流组织和检索包结构及各自功能。本课题索引存储结构采用倒排表结构,有效的存取倒排表和快速响应搜索操作在全文搜索领域中起着至关重要的作用。本文对索引存储结构进行了深入的分析,设计一种能够有效提高索引压缩率的索引文档重排算法——聚类索引重排算法。该算法将相似度高的文档排列在一起,从而减少编码文档编号之间的差值所需要的字节数,达到提高搜索效率的效果。通过测试,聚类索引重排算法可以明显减少索引存储空间,达到索引压缩效果更多还原

【Abstract】 The rapid development of information technologies and the advent of the Wore-Wide Web have resulted in a tremendous increase in the amount of available information, especially the data in the format of doc、pdf and txt are also expansion rapidly. Facing the more and more popular trend of the network electronic information, how to reasonable arrangement the format and content of the information and how to let them to bring the biggest convenience to the internet users, this is the research goal of this thesis. The research aim of this subject is to build a chorography Full-Text retrieval system based on B/S which accessed via web. The research target of this subject is to design and develop a Chinese full-text retrieval system, thus realizing the full-text search functionality of local gazetteers. An improved inverted index algorithm named Cluster index realign algorithm is designed and used to store the compressed mass index file, for reducing the index storage space and improve retrieval efficiency results.The subject studied the principal of full-text retrieval in details, including the principle of full-text retrieval index, index content, indexing and search. This paper analyses the structure, index process, search process, the data flow, organization structure and their respective bag and retrieval functions of Lucene.The index storage structure has been analyzed further in this paper, designing an algorithm which could effectively improve the compression ratio of index----cluster index realign algorithm. This algorithm array the high similar document together in order to decrease the small d-gap frequency through clustering, which would make the index compression more effective.Accroding test, cluster index realign algorithm could decrease the storage space of index,and at last, accieve index compression effect.更多还原