There are not many texts covering the whole range of database design from requirements analysis to physical design. So this book deserves appropriate attention by the academic community as well as practitioners in the field.
The 18 chapters of the book present the material in the order that a normal database design process is supposed to follow. The chapters are grouped into five sections called introduction, conceptual design, implementation design, physical design, and special design issues.
The introductory section gives an overview of database systems, the basis steps of the database design process, and requirements formulation and analysis as the first step in design, the others corresponding to the next three sections of the book as mentioned above. The major design issues considered are integrity, consistency, recovery, security, efficiency, and the effects of projected future growth.
Conceptual design is treated in three chapters dealing with conceptual data modelling, entity formulation and analysis, and attribute synthesis. For conceptual design, two perspectives are discussed, object modelling involving abstraction, aggregation, generalization and user views, and entity modeling involving entities, attributes and relationships. Much attention is given to methodologies for conceptual design, including entity analysis as a top-down approach and attribute synthesis as a bottom-up approach.
The major objective of implementation design is to produce a DBMS processible schema that satisfies the requirements established in previous steps. The design process for schemas and subschemas and the detailed efficiency analysis given is mainly oriented towards CODASYL- like network databases. There is no material on the peculiarities of relational database design.
The largest section, containing nearly half of the material of the book, is devoted to physical design involving record structure design, record clustering, primary and secondary access methods, and secondary index selection. The material is treated extensively in eight chapters, as much as the first three sections together.
The special design issues discussed in the final section are reorganization and distributed database design. There are three appendices containing useful exercises for the design steps and a list of variables used in the book. A glossary, a rich list of references, and an index contribute to the usefulness of the book.
The book establishes a consistent framework for multilevel database design, defines a workable methodology, and describes a set of general principles, tools, and techniques for database design at each level. Useful examples show how the methodology can be performed manually, but they also point out where computer-based tools can be implemented. The scientific foundation of physical design aspects is much deeper and firmer than that of conceptual and implementation design concepts. The heavy emphasis on physical design in this book reflects this state of the art, but it is somewhat in contrast with the tendency of modern DBMS to free the database designer from more and more parts of physical design. The perhaps most serious point of criticism is the bias towards network databases, omitting all aspects of relational database design.