Information Integration Based On Federated Database

With the development of informationization and the exploresive increase of the available information of human society, Information Integration becomes an urgent requirement. Usually, the requirement of Information Integration is due to some global demands, which requires data integration of distributed, heterogenerous and automomous information systems using universal application framework. In many cases, the objects of Information Integration are bequeathed databases which mean existing databases. Federated Database is an effective method of the Information Integration of bequeathed databases.Distribution, heterogeneity, autonomy and transparency are the most important characteristics of Federated Database System. This article explores the heterogeneity, especially the semantic heterogeneity, which is divided into two categories, one is structural semantic heterogeneity which will be solved without compromising the autonomy of component databases, the other one is fundamental semantic heterogeneity which will compromise the autonomy. In order to achieve the four characteristics, the solution given in this article chooses the tightly-coupled federated databases with one federated schema to support the transparency; Chooses B/S 3-tier structure to make the unique entrance and convinient access, and supports the distribution of component databases using Internet; Shields the heterogeneity of component databases in techniques and category making use of wrappers; Provides the possibility of the integration of component databases with semantic heterogeneity through 5-tier architecture of schema. Metadata plays an important part in schema integration. A Metadata Object Model denoted in a sextet is advanced, and a Metadata Object Diagram is developed from Metadata Object Model to distinguish fundamental semantic heterogeneity and benefit the creation of federated schema and the process of queries in federation. Metadata Object Model can express the metadata of both the related databases and the XML documents which are the representatives of semi-structured data. It provides a carrier for both of the two kinds of data. Based on the Metadata Object Model, this article proposes the methods of schema integration and process of queries in federation putting emphasis on the solution of structural semantic heterogeneity, and puts forward the query method of XML data in federation.