A Study On Heterogeneous Data Integration Based On XML And Its Implementation

An approach to set up the conversion between the XML schema and the relation schema in the progress of isomerous data integration is presented in this dissertation. In this approach, the conversion rules between XML schema and relation schema are defined first. Based on the XML schema document tree, a node tree is created. Then the conversion is implemented according the node tree. It not only reflects the structure characteristic of the data, but also keeps most of the semantic attributes and constraints of the data. After that, in order to solve the the problem that how to store XML document into database, a relation schema of middle-size granular is defined under the Dietz region-based coding schema. In this method, the XML document was broken into fragments first, and then stored them into database. At last, for the different size XML document, the same experiment has been done. Some data were collected, and analyses were done at the same time. According to the analyses, the middle-size granular method can save more time than the little-size granular method.