Research Of Multidimensional Metadata Index And Query On Logistics Information Files

With the rapid development of electronic commerce and logistics industry, the logistics files have many characteristics, such as frequent updates, large amount of data, in logistics storage processing system. It s a urgently solved problem that how to quickly find the logistics files. Logistics files contain large number of multidimensional metadata. The existing index structure can t effectively meet the requirements of the multidimensional metadata complex queries, because it did not make full use of the metadata. According to the above questions, the main research is as follows:First of all, according to the demand of the multidimensional metadata index structure in the logistics files. It is presented that a multidimensional metadata index structure, which is put bloom filter into R-tree nodes. So, it is called R-tree and bloom filter, hereinafter referred to as R-TBF. At the same time, because of metadata update frequency frequently in logistics files, the update mechanism of multidimensional metadata index structure is proposed, to ensure the index structure update performance. The update mechanism is composed of basic index structure and update cache structure. Multidimensional metadata query instances is used to explain the multidimensional metadata index structure and the update mechanism of multidimensional metadata index structure.Secondly, in order to quickly query the logistics files from the logistics information storage system, the query mechanism of R-tree and bloom filter based metadata cube is presented. It’s called for short R-TBFMC. The query mechanism designed three modules: main module, query module and update module. In addition, two kinds of query methods are proposed, which are range-sum query and range-max query.Finally, according to the design idea of the multidimensional metadata query mechanism, three modules of the query mechanism are accomplished. The range-sum query and range-max query are also finished. At last, the experiment and analysis about the performance of the query mechanism is completed, which show that the performance of the query mechanism is high-efficiency.