XML Data Management Based On Relational Database

Due to adaptable structure and flexibility in defining the tags, XML has recently emerged asthe leading medium for data storage and data transfer over the World Wide Web. Manyorganizations had adopted XML as the principal facet in their online business applications. Onthe other hand, relational database is still widely used as the back-end database in mostorganizations. Thereby, how to effectively map XML documents to relational database becomesa hot topic for current researchers. This paper proposes an effective mapping scheme to bridgeXML and relational database based on the model-mapping method. Our approach has only threepre-defined tables to store the structural information of leaf node, paths of leaf nodes and innernodes within the tree structure respectively, which save the storage size. On the other hand, thecombination of path indexing and numbering label supports the representation ofancestor-descendant and sibling relationship between nodes.The relational schema is used for an efficient processing of queries in particular queriesinvolving structural conditions. The chain and unchained path expression can be converted to thecorresponding SQL statement while avoiding redundant joins. And the mechanism also putsforward an algorithm of conversion for handling the traversal query in the part of XMLdocument.Finally, An experiment based on an middleware prototype is implemented forverifying the mapping and converting algorithm with the collection XML document of differentsize and complexity. The test data indicated that our strategy and algorithm is robust in terms ofdatabase storage, data loading and query time of complex queries.