A data warehouse is an integrated database primarily used in organizational decision making. Although the deployment of data warehouses is current practise in modern information technology landscapes, the methodical schema design for such databases has only been studied cursorily.
This book presents conceptual and logical schema design phases as parts of a general design process for data warehouses. The specification of conceptual schemata and the assessment of their quality is based on a formal multidimensional data model in conjunction with assiciated normal forms. Moreover, a conceptual design phase is developed whose output schemata provably satisfy the normal forms. Logical design is supported by an algorithmic approch, which transforms conceptual schemata into relational database schemata that satisfy several desirable properties. Furthermore, the notion of self-maintainability is generalized towards independance, and an approach towards warehouse independance based on so-called view complements is proposed.