Nattapong Thawornkool. A table-based clustering strategy for XML data storage and querying. Master's Degree(Computer Science). Mahidol University. : Mahidol University, 2009-04-22.
A table-based clustering strategy for XML data storage and querying
Abstract:
XML, a data format, has become the de facto standard for interchanging data among sources. As the processing of XML data is increasing, the requirement of efficient XML data storage and querying is apparent. Native and non-native XML storage are two aspects of storing XML data that facilitate efficient XML query processing. The element-based and the subtree-based clustering strategies are used to cluster XML data in the native storage with respect to the type of XML data elements and the sub-tree structure, respectively. In this thesis, a new table-based clustering strategy to support efficient native storage and querying of XML data is described. The strategy combines element-based clustering and the subtree-based clustering. It categorizes the XML data with respect to the type of data elements and classifies the tree-like structure of XML data regarding the cardinalities of relationships between the data elements. The clustering technique stores XML data into a set of tables. With table-based clustering, the size of XML files can be reduced by half. Additionally, table-based clustering has been evaluated with a good performance for complex queries. "