Xml Heterogeneous Data Integration Technology-based Applications

In the Digital Campus process, the heterogeneous data in the independent information systems needs to integrate together because of the data is unable to communicate and share. For realize the data integration through set up a unified data exchange platform, the public data query interface which can mask the diversity owing to different data sources was required. Therefore the XML has now become the standard for data exchange, in virtue of it's characteristic such as being independent of the platform, easy to expand and rich in meaning to describe the data. And XML-based data integration technology was a hotspot in this research field.In this paper, the XML/middleware technology and other related research areas is the research background and the project of Digital Campus is the research environment. On the base of analyzing and studying the key technologies of heterogeneous data integration, combined with the state of constructing the digital campus, the paper designed a Heterogeneous Data Resource Integration Middleware(HDMI)based on the XML. In view of the significance of query decomposition and optimization in the Heterogeneous Data Resource Integration System, in-depth research on this aspect was mainly conducted.1. By researching on the XML, middleware and data resource integration technology, the HDMI used the Wrapper/Mediator method. The paper introduces the architecture of the HDMI, analyses and designs the Functional modules. Middleware hadn't the specific data but memorized the integration model is the main characteristic of this scheme. HDMI system supplies each data resources with a specific interface. Wrapper is used for packing the heterogonous data resources. Mediator parses and deals with the user layer's request and return the result to the user layer.2. The paper introduces the steps of query process, gives the particular way of the query decomposition, and also analyses the query optimize method. Query decomposition focused mainly on the FLWR expressions. According to the FLWR expression characteristics, the algorithm is divided into three parts: decomposing the LET and FOR clause, handling RETURN clause and WHERE condition. Query Optimization mainly considered on four aspects: optimizing the query decomposition, optimizing the system, optimizing the query plan, analyzing the impact of network transmission to the query operation.