Browse Books

Reviewer: David Gary Hill

The world of information technology is often full of hyperbole, so the use of the words "father," "Bible," and "classic" in the same paragraph may seem excessive. However, they are appropriate in the context of this book. Data warehousing is now a very important part of information technology, and credit for being the father of the field is an appropriate designation for the author. This standard data warehousing reference book has long been a comprehensive introduction to the core concepts and methods of data warehousing. Therefore, the terms "classic" and "Bible" apply to it. As befits a fourth edition, the book builds upon its previous base with discussions of newer topics, such as compliance and multidimensional database design. The book can serve as a reference guide to remind experienced data warehousing professionals of the first principles of data warehousing. However, information technology professionals who are new to data warehousing can use the book to immerse themselves in understanding the fundamentals of data warehousing. The topics in the book are remarkably easy to follow and easy to absorb. Numerous simple, clear figures illuminate the discussion. Most of the first half of the book delves into the aspects of environment, design, and technology that are necessary for the data warehouse. Fundamental to understanding the data warehouse environment is knowing that the system development life cycle of the data warehouse is essentially opposite to the standard waterfall system development life cycle. A data warehouse follows a "build it and they will come" methodology, where requirements are specified after the data warehouse is built. The key characteristics of a data warehouse are that it is subject-oriented, integrated, nonvolatile, and time-oriented. The purpose of a data warehouse is to support decision making, and this means that it has to be able to support requirements that have not yet been defined. The chapter on data warehouse design goes on to describe the data model that is necessary for a data warehouse, especially since the life cycle methodology says that a data warehouse has to be built iteratively. A key decision for a data warehouse is the level of granularity of the data. If the level of detail is too voluminous, the performance of the data warehouse can be seriously affected. This does not mean that the data warehouse will not have to manage a large volume of data. The technology chapter discusses managing a large data warehouse efficiently. The chapter also discusses multidimensional database management system processing (also known as online analytical processing). The book then goes into a number of architectural issues, including distributed data warehouses, integrating external data into the data warehouse, how to work with the Web, and how to deal with a really large data warehouse. The book covers the relationship of a data warehouse with data marts and operational data stores. The business issues of cost justification and return on investment, as well as corporate information governance, each receive chapter-length treatment. The book concludes with an extensive data warehouse design review checklist. Some key topics, such as granularity, are covered in more than one place in the book. Some readers may not like this redundancy, but the repetition should help with the retention and reinforcement of key points. Finally, the book contains a useful glossary and an extensive list of references. Overall, the book covers the subject of data warehousing thoroughly. The world of information technology is often full of hyperbole, so the use of the words "father," "Bible," and "classic" in the same paragraph may seem excessive. However, they are appropriate in the context of this book. Data warehousing is now a very important part of information technology, and credit for being the father of the field is an appropriate designation for the author. This standard data warehousing reference book has long been a comprehensive introduction to the core concepts and methods of data warehousing. Therefore, the terms "classic" and "Bible" apply to it. As befits a fourth edition, the book builds upon its previous base with discussions of newer topics, such as compliance and multidimensional database design. The book can serve as a reference guide to remind experienced data warehousing professionals of the first principles of data warehousing. However, information technology professionals who are new to data warehousing can use the book to immerse themselves in understanding the fundamentals of data warehousing. The topics in the book are remarkably easy to follow and easy to absorb. Numerous simple, clear figures illuminate the discussion. Most of the first half of the book delves into the aspects of environment, design, and technology that are necessary for the data warehouse. Fundamental to understanding the data warehouse environment is knowing that the system development life cycle of the data warehouse is essentially opposite to the standard waterfall system development life cycle. A data warehouse follows a "build it and they will come" methodology, where requirements are specified after the data warehouse is built. The key characteristics of a data warehouse are that it is subject-oriented, integrated, nonvolatile, and time-oriented. The purpose of a data warehouse is to support decision making, and this means that it has to be able to support requirements that have not yet been defined. The chapter on data warehouse design goes on to describe the data model that is necessary for a data warehouse, especially since the life cycle methodology says that a data warehouse has to be built iteratively. A key decision for a data warehouse is the level of granularity of the data. If the level of detail is too voluminous, the performance of the data warehouse can be seriously affected. This does not mean that the data warehouse will not have to manage a large volume of data. The technology chapter discusses managing a large data warehouse efficiently. The chapter also discusses multidimensional database management system processing (also known as online analytical processing). The book then goes into a number of architectural issues, including distributed data warehouses, integrating external data into the data warehouse, how to work with the Web, and how to deal with a really large data warehouse. The book covers the relationship of a data warehouse with data marts and operational data stores. The business issues of cost justification and return on investment, as well as corporate information governance, each receive chapter-length treatment. The book concludes with an extensive data warehouse design review checklist. Some key topics, such as granularity, are covered in more than one place in the book. Some readers may not like this redundancy, but the repetition should help with the retention and reinforcement of key points. Finally, the book contains a useful glossary and an extensive list of references. Overall, the book covers the subject of data warehousing thoroughly. Online Computing Reviews Service