基于多粒度索引的互联网信息检索系统

摘要

Based on the technologies of Internet information retrieval ,the multi-granularity indexing technology which utilizes the statistical method to combine the phrases index and words index is proposed .This method first identifies the un-known words contained in the document and then puts these unregistered words in an extended dictionary .When the index is created ,word segmentation for the text is conducted tuice .Based on multi-granularity indexing technology ,Internet infor-mation retrieval system is designed .Experimental results show that the system has achieved good results in accuracy ,recall-ing rate and composite score F-measure .%总结了互联网信息检索的相关技术，提出了多粒度索引的技术。多粒度索引技术采用了统计学方法，综合了短语索引和词语索引的优秀思想。该方法先识别文档中包含的未登录词语，然后把识别出来的未登录词语依次放在一个扩展的词典里面。在创建索引的时候，对文本进行两趟词语切分。第一趟是采用基本的分词词典。第二趟词语切分是对上面的结果采用基于扩展分词词典的方法。基于多粒度索引技术设计了互联网信息检索系统，实验数据显示系统在准确率、召回率以及综合评分F-measure上均取得了较好的效果。