物流信息文件多维元数据索引与查询研究

发布时间：2018-10-17 17:41

【摘要】：随着电子商务和物流产业的迅猛发展,其存储处理系统中的物流文件具有数据量大、数据更新频繁等特点,对物流文件如何快速地实现查找是物流系统迫切需要解决问题之一。物流文件中含有大量的多维元数据,现有的索引结构因没有充分利用其中元数据,而不能更有效的满足多维元数据复杂查询的要求。为此,本论文主要研究内容如下:首先,在物流文件中,根据多维元数据索引结构的需求,本文在R树节点上放入布隆过滤器(Bloom Filter)提出一种多维元数据索引结构,称为R树布隆过滤器,简称R-TBF(R-Tree and Bloom Filter)。物流文件中元数据更新频次高,为了保证索引结构的更新性能,本文接着提出R-TBF索引结构的更新机制,它由基础索引结构和更新缓存结构组成。最后用在物流文件中的多维元数据的查询实例说明我提出的索引结构与更新机制。其次,为了能够在物流信息存储系统中快速的查找到物流文件,本文采用R-TBF索引结构,建立元数据立方体(metadata cube),设计了一种基于元数据立方体的R树布隆过滤器的查询机制,英文简称R-TBFMC(Query Mechanism of R-Tree and Bloom Filter based Metadata Cube)。查询机制设计了三个模块:R-TBFMC主模块,R-TBFMC查询模块以及R-TBFMC更新模块。同时设计了两种查询方法,分别是范围和值查询法和范围最大值查询法。最后,进行了多维元数据查询机制的实现工作,实现了多维元数据查询机制的三个模块,以及两种查询算法——范围和值查询法和范围最大值查询法。并对提出的查询机制(基于元数据立方体的R树布隆过滤器的查询机制)的性能进行测试与分析来证实此查询机制的高效性。
[Abstract]:With the rapid development of electronic commerce and logistics industry, the logistics files in the storage and processing system are characterized by large amount of data and frequent updating of data. How to find logistics files quickly is one of the urgent problems that need to be solved in logistics system. There are a lot of multidimensional metadata in the logistics file, and the existing index structure can not satisfy the requirement of complex query of multidimensional metadata because it does not make full use of the metadata. Therefore, the main contents of this paper are as follows: firstly, according to the requirement of multidimensional metadata index structure in logistics file, this paper puts a kind of multidimensional metadata index structure (Bloom Filter) into the R-tree node. Called R-Tree Bron filter, abbreviated as R-TBF (R-Tree and Bloom Filter).) In order to ensure the update performance of index structure, this paper proposes an update mechanism of R-TBF index structure, which consists of basic index structure and update cache structure. Finally, the index structure and update mechanism are illustrated by a query example of multidimensional metadata in the logistics file. Secondly, in order to find the logistics files quickly in the logistics information storage system, this paper adopts the R-TBF index structure, establishes the metadata cube (metadata cube), to design a query mechanism of R-tree Bron filter based on the metadata cube. Abbreviated R-TBFMC (Query Mechanism of R-Tree and Bloom Filter based Metadata Cube). The query mechanism includes three modules: R-TBFMC main module, R-TBFMC query module and R-TBFMC update module. At the same time, two query methods, range and value query and range maximum query, are designed. Finally, the implementation of multidimensional metadata query mechanism is carried out, and the three modules of multidimensional metadata query mechanism are implemented, as well as two query algorithms: range and value query and range maximum query. The performance of the proposed query mechanism (the query mechanism of R-tree Bron filter based on metadata cube) is tested and analyzed to verify the efficiency of the query mechanism.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP311.13

【相似文献】