Database management systems (DBMSs) have greatly influenced the development of successful commercial systems. Generally, the DBMS environment is suitable for transaction-oriented systems. The concept of decision support systems (DSS) is recent, introduced by Keen and  Morton [1].  It was given a formal structure by Sprague [2]. Recently, a large number of papers based on applications using DSS concepts have been published.

This book is concerned with the development of DSS applications using appropriate databases. The author claims that the methodology of design of a database may not be suitable for DSSs. He has introduced an important concept, the data warehouse; this book mainly deals with the design of the data warehouse.

The book is divided into ten chapters followed by an exhaustive appendix. Chapter 1 introduces the concept of a data warehouse. The main goal of the chapter is to emphasize the difference between the classical operational development life cycle and DSS application development. The classical development cycle consists of requirements, database design, and program development. According to the author, data warehouse design starts with the design of the data warehouse, followed by the development of programs and then understanding requirements. Inmon also introduces the concepts of operational data, data warehouse departmental data, and individual data.

In chapters 2, 3, and 4, the important design considerations of a data warehouse are identified as granularity and partitioning. These two aspects are used to explain the design of a data warehouse in later chapters. A data warehouse involves a large volume of corporate data. Hence granularity, based on the level of detail, and partitioning of data are important.

Data warehouse design depends greatly upon the available technology, such as language interfaces, composite keys, and variable-length data. A distributed data warehouse consists of local data warehouses and global data warehouses. Chapters 5 and 6 give details about these concepts.

Chapter 7 considers executive information systems (EISs), a specific type of DSS, in the context of the data warehouse. A data warehouse has to hold external data as well as the corporate data for DSS applications. Such external data may have a different structure from that of the internal data or they may be unstructured. Chapter 8 introduces metadata and its role in acquiring the external data.

The basic strategy of implementation of a data warehouse is incremental. Migration from the present setup to the data warehouse is the major topic of the last two chapters. Design review is an important aspect of data warehouse design. The author emphasizes the differences between operational system design and data warehouse design. A large number of review considerations--54, to be exact--are discussed. This checklist is helpful for the designer of a data warehouse.

The 50-page appendix is useful. It integrates the design of operational data and the data warehouse through details of steps and deliverables. A list of important references and a glossary conclude the book.

Most of the concepts of systems analysis and design are simple and based on observation and practice. The book would have been more useful if a complete case study had been included. The reader would have gotten more details about the design. I strongly recommend this book to designers of DSSs, EISs, and other variations of these concepts. It is practical and has sufficient details to guide the designer toward development of a good data warehouse. This book can also serve as a supplementary reference book for a DSS course for graduate students, though the lack of examples at the ends of chapters may be a limitation. It is a definite contribution to the growing and important field of DSS.