This thesis describes a method of integrating a small number of separately developed heterogeneous databases and permitting the participation of a new database. The composite and federated approaches are combined in establishing a hybrid approach to database integration and resource sharing with a global schema and two query language translators from SQL to OPAL or vice versa. The approach overcomes incompatibilities in syntax and semantics of queries by using two mapping tables and a parser table to enable them to be accessed and modified coherently. Lexical analysis, syntax analysis and semantic analysis are used in translating a query in one language into the other. The major advantages of the hybrid approach are simplicity, effectiveness, and efficiency while it is used to integrate a small number of heterogeneous databases. Evaluation of schema integration and semantic matching is based on three different types of data models (relational, entity-relational and object-oriented databases) with similar data in semantics. The basic ideas are illustrated using a few examples. A prototype system is implemented using the TCP/IP protocol in a C/Unix environment. |