Study And Implementation Of A Heterogeneous Data Integration System Based On XML

With the rapid development of Internet, sharable resources on the network areincreasing and the descriptions of data differ from each other in a thousand ways. Howto integrate data from these distributed, heterogeneous and independent data sources, aswell as to ensure the integrity and consistency of the data becomes a very importantresearch topic.Against this problem, this paper provides a multi-source heterogeneous dataintegration system. This system uses XML as data exchange format, combined withMediator/Wrapper architecture and semantic caching to achieve the query optimization.In the beginning, this paper introduces the related techniques about heterogeneous dataintegration. What’s more, the entire research work is presented specifically. Firstly, byanalyzing existing approaches of data integration, a scalable heterogeneous databaseintegration system is proposed. Followed by the system structure view as well asanalysis of each module. Secondly, using XML as data exchange format, the local datasource is transformed into XML data for integration. A unified and transparent accessinterface is provided by shielding the underlying heterogeneous data sources. Thirdly,using XQuery as a query language on the global schema, in order to decompose thequery and convert into SQL statements. Fourthly, by semantic query caching, the queryis optimized and the response time is faster. At last, various data sources are wrappedinto Web services which eliminate the differences between data sources, makes theintegrated system loosely coupled and easily extensible.Finally, the proposed system is tested through experiments. The experimentalresults demonstrate the feasibility and correctness of this integrated system.