Research On The Xml-based Management Of Heterogeneous Data Sources

Research on The XML-based Management of Heterogeneous Data SourcesIn this paper, the solutions of XML-based integration of heterogeneous database middleware are proposed, which is the main face of heterogeneity, integrity, performance, semantic conflicts, permissions bottleneck, additional constraints, and integrated content, when build heterogeneous database integration system. Through the implementation of transformation and integration relational data to XML, the middleware provides support for sharing data, and application of integrated access to the underlying information. The main contents are as follows:1. Established connection between the heterogeneous databases. Only establishes a connection between the conditions of heterogeneous databases can provide a global query to the users. A key to integrate and connect the data which is stored in heterogeneous data sources, is to determine the correspondence between the semantics of the data source. The correspondence also exists in the schema and instances level. Therefore, cluster analysis techniques used in this paper, in a synergistic way to identify and analysis the mode of correspondence between schema and instance-level correspondence. Proposed a probability technology based on logistic regression estimation method, which can more effectively identify entities match in heterogeneous environment.2. Mask the differences between heterogeneous databases, and provide a global view to the users. The purpose of this project is to use software to simulate the physical database, to provide users a global view of heterogeneous distributed databases, enabling the users to focus more on business without concern for details of the underlying database. Using the characteristics of heterogeneous integration middleware, propose and design mediator-wrapper architecture to contract the necessary for business processes, applied underlying services. Build a XML-based type system and a Java-based data type conversion system. Using XML technology mask the differences databases; provide a good description of the relationship between the method for the database; XML file is a text file, with good interchangeability, not limitation by the operating system and software platform, currently, numbers of database provide support for XML. 3. There are two modules in this project: the global query module and the data source created module. The global query module is based on Native XML data type system, providing underlying database and a global view of the database for the user. The module create the thinking of object-oriented, offer the interface between different types of database to convert data; introduction of DOM, SAX technology to achieve global query over XML document of heterogeneous databases. Build a module in the data source to provide the user a new way that use existing data sources to create a database, and allowed the user to build a new database content from existing database. The project using XQuery inquires XML data, and proposes an efficient XML document storage and query mechanisms; defines a database for the heterogeneous distributed query language. The character of this query language is easily controlled by customer, with simple clear statement and algorithm, improved the query efficiency, and helping users to easily create a new physical database and import data from existing data sources.