Research And Realization For Heterogeneous Database Integration Middleware Based On XML

With the rapid growth of information, more and more applications have to access to heterogeneous data sources. Due to the geographical distribution of these different data sources with different platforms, different operating systems, different database models, different semantic representation, and different database management systems (DBMS), it is always difficult to access these data. So, it is urgent to use existing information technology to dynamically integrate these heterogeneous data resources and make information sharing efficient.Data integration is to meet the needs of different businesses, to realize the sharing of heterogeneous data sources and data transparent access. At present, there are many solutions in the data integration. However, there solutions are always complicated and subject-oriented. Although the traditional database middleware seem more flexible, it is difficult for the integration of information systems because that tightly coupled application model allows less independent middleware, deployment and use more difficult. How to build a database middleware in an efficient, general-purpose, scalable, cost-effective way will be an important direction in the field of database middleware.According to the edge of XML in the data integration and exchange, the paper resents a XML-based heterogeneous database integration middleware. We firstly introduce the knowledge of heterogeneous data integration, analysis the data integration technologies and methods of existing emphatically, and compare their advantages and disadvantages. Through XML integration model, we elaborate emphatically the design ideas and the various functional blocks of the implementation process of the XML-based heterogeneous database integration middleware. Through the realization of mapping, transformation and integration from local model to the global view, this middleware shields all differences in aspects of data source including the platform, system environment and internal data structure, and so on. It provides a unified and transparent interface for users to implement the accession. Meanwhile, The key point of the research is how to better address cross query, joint query and other complex queries between the various sub database in the environment of distributed heterogeneous. To this end, after discussing the general method of the centralized database and distributed query, this paper resents a query decomposition algorithm based on the global virtual view, which can solve the related query between multi-databases. Because of involved a large number of data transmission in the integrated system, and in the environment of limited network resources, how to reduce the amounts of data transmission and query response time must be considered. Based on the semi-join query optimization algorithm, this paper improved the algorithm. Through the analysis of qualitative and quantitative aspects, the improved semi-join algorithm has fewer amounts of data transmission and response time than the general semi-join algorithm, and the performance is more superior.Finally, this paper implements the system by experiment of the integration of XML-based middleware. Through the application of that, the XML-based integration middleware can better integrate heterogeneous data, implement the sharing of resources, and provide data access efficiently and transparently.