Research And Application On XML-based Heterogeneous Data Sources Integration System

With the development of network and rapid increasing of data on the Internet ,it is necessary to build a data integration system to share the entire heterogeneous information source.The function of a data integration system is to provide a uniform interface to a multitude of data sources. Using the interface users don't need to consider the problem about model heterogeneous, data extracting and combination, etc..The problem of data integration is one of the old classical problems in database community. But with the rising of XML technology, the problem again becomes a research focus in this community.XML has emerged as the standard of meta language representing data format, and is continually applied to various domain to integrate data sources. It is self-description,open extensible and platform independency. All the advantages make it the best candidate for representing data model.Combined with the advantage of XML ,this paper discusses some theories and technologies of heterogeneous data integrating. Based on analysis and comparing the existent data integration method and the character of architecture we design a HDSISBX system (heterogeneous data sources integration system based on XML). Meanwhile, we discuss the main technology. The main contribution of this paper is as follows:1. Aim at the lack of the traditional data integration system architecture, this paper analyses and compares the existent data integration method and the character of architecture.Considering the characteristic of the domain data ,we ameliorate the current architectures and design a heterogeneous data sources integration system architecture based on XML.This paper gives the realized strategies of all modules and implements some key modules of the system.2. Research on pivotal technology :1) Main Schema Extractive: Combined with the advantage of XML ,this paper gives a XSDM model(XML Schema-based Data Model) and regards it as the common data model to describe the schemas of all the data sources.Main schema extractive extractives the key schema information to depict the application topic from the data sources based on the unification of all the data sources schemas.The system builds a main schema collection composed of the key schema informations, it synthesizes the schema information on the entire domain application.The integration system takes it for unified view to query domain heterogeneous data sources for user. This paper analyses the main steps of the main schema extractive process,gives the arithmetic and maintenance strategy of the main schema extractive.2) Query processing: This paper presents the query capability of the data source, analyses the query decomposed process based on the query capability and gives it arithmetic.