基于VSM的OAI-PMH元数据相似度计算研究  

Research of Calculating Metadata Similarity in OAI Framework Based on VSM

作　　者：赵治军[1] 陈立潮[1] 谢斌红[1] 王秀慧[1] 

机构地区：[1]太原科技大学计算机科学与技术学院,山西太原030024

出　　处：《计算机技术与发展》2009年第9期119-121,199,共4页Computer Technology and Development

基　　金：太原市科技项目(ZB0701-04);校青年基金(2007133)

摘　　要：针对OAI-PMH存在大量相似的元数据,结合元数据内容的结构特性,提出了一种基于向量空间模型的OAI-PMH元数据相似度的计算方法。为了较好地反映特征项在元数据内容不同层次的重要程度,采用了结构层次权重系数改进的TF-IDF方法来计算特征项权重。实验分析表明,使用该方法对元数据之间相似度的计算是有效的、可行的。为用户在搜索查询时提供了基于元数据的相似资源,方便了用户,提高了信息服务质量。A large number of similar metadatas exist in the OAI- PMH. In order to calculate the similarity among metadatas, a method which is based on VSM is proposed while considering the structural characteristic of metadata content. In order to reflect the importance of characteristic term in different layer of metadata content, the improved TF- IDF which is based on structure - layer weight coefficient is used to measure the characteristic term in VSM. The experimental analysis shows that the use of the method for calculating the similarity among metadatas is effective and feasible. It provides metadatas for users which is based on similar resources,facilitates users and improves the service quality of information.

关 键 词：元数据相似度 向量空间模型 结构层次权重系数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]