A methodology using data base abstractions for logical data base design has been proposed. This paper describes the application of that approach to a data base design problem, and evaluates the methodology with respect to three criteria: ease of use, completeness, and correctness.

The decisions which must be made during the implementation of a very large data base may be classified into three major categories: decisions related to data base organization, those concerning the selection and use of data storage devices, and those related to data compression. The logical organization of the data base must be determined to suit data base users and a corresponding physical organization must be selected to implement that logical organization in an efficient manner. The type and number (size) of storage devices to be used must be determined. The nature of these devices and whether the storage bank will be homogeneous or consist of a mix of device types must also be determined, Finally, the feasibility of data compression must be investigated. If desirable, an efficient method must be selected and the extent of its application determined.