Data sharing economics and requirements for integration tool design

摘要

In most of today's companies we find heterogeneous database systems containing redundant and inconsistent data. This threatens the ability to make coordinated, organization-wide responses to business problems. Benefits of data integration need not dominate the costs. In spite of this, we argue that some kind of common understanding of information structures is necessary. Thus, data integration is necessary at least to some degree. In the literature only technical aspects of schema integration are treated. Also, the complexity issue (large schemata are hard to understand) is usually not addressed explicitly. We present a business process oriented strategy for data integration. This method allows the determination of the order and the degree of integration. Complexity is reduced by schema clustering during the pre-integration phase.