一种改进BIR模型及其在FTP搜索中的应用  

Modified BIR model and its application in FTP file retrieval

作　　者：吴以才[1] 刘渊[2] 郑楼英[3] 

机构地区：[1]浙江大学信息学院,浙江杭州310027 [2]浙江大学管理学院,浙江杭州310027 [3]浙江经济职业技术学院,浙江杭州310012

出　　处：《浙江大学学报（工学版）》2001年第5期563-566,共4页Journal of Zhejiang University：Engineering Science

摘　　要：探讨把文档检索中的二值独立检索模型改进后应用于 FTP文件检索 .解决现有仅通过字符串匹配检索时存在的无法按相关程度排序结果和因文件名采用缩写时造成漏检的问题 .检索模型的参数估计采用启发式规则自动获得训练集 ,实现检索全过程自动化 .实验表明 ,使用该模型的检索结果比通过字符串匹配法具有较高的精确度和检全率 .A modified BIR model was developed for FTP file retrieval. It solved the problems happened to current retrieval method based on string match, which can't rank the retrieval results and may miss retrieval targets when filenames are abbreviated. Some heuristic rules were proposed to obtain training set for estimating the parameters of our retrieval model, and enabled the new model to do the whole retrieval process automatically. The experiment showed that the results of the new retrieval model have higher precision and recall than the string-match methods'.

关 键 词：FTP检索 文档检索 二值独立检索模型 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]