Understanding Structural and Semantic Heterogeneity in the Context of Database Schema Integration

Over the last 25 years data integration has been a key issue in achieving systems interoperability, between heterogeneous data storage and management systems, because of the existence of system, schema, and semantic heterogeneity. This paper starts by examining the background to data integration by discussing why integration has remained such an issue, in terms of current rationale and drivers for data integration. Previous integration strategies are briefly discussed and these encompass three generations of methodologies. They address multiple schema integrations, using federated databases and mediator/wrapper systems and, more recently, the trend towards information brokering using ontology-based approaches. However, the integration process primarily requires understanding of the characteristics of heterogeneity, so that database schemas can be examined to identify semantically related conflicts that may then be transformed, or mediated, to deliver interoperable data stores. Therefore, the more detailed consideration in this paper is towards an understanding of both the causes and examples of structural (schema), syntactic (format) and semantic (meaning) heterogeneity, created during the data model and schema design process. Finally, the Semantic Web is introduced because associated technologies may provide the language specification to underpin third generation approaches of ontology development.