Customer Reviews

Object-Oriented Data Warehouse Design: Building A Star Schema

5 star:		(5)
4 star:		(1)
3 star:		(1)
2 star:		(0)
1 star:		(2)

 

 

If you are just playing at data warehousing, then get yourself a copy of one of those "The Doofus' Guide To Whatever" books. But, if you want to successfully implement an enterprise-level, mission-critical data warehouse, this is one book you will definitely want to have. "Object-Oriented Data Warehouse Design: Building A Star Schema" presented me with a new look at the whole Business Intelligence field. I am not even half way through and I can already see why some past projects ran into trouble and how I can avoid similar problems in the future.

And this is not one of those unreadable academic tomes that you put on your bookshelf to impress your boss. I have to agree with Bill Inmon's foreword that Giovinazzo "strikes a fine balance between theory and practicality. Theories are explained in the cloth of practicality. Rules of thumb and practical realities always have a touch of theory to explain the underlying philosophy." This is a very readable book with lots of immediately useful information without resorting to that "cookbook" approach. I especially enjoyed the discussion in Chapter 4, The Implementation Model, on multiple dimensions and translating them into a star schema. The whole object oriented approach to star schema design seems so logical, now that I have had it explained in such a clear, concise manner.

BTW, the other book I wouldn't be without is Ralph Kimball's "The Data Warehouse Toolkit". But if you can only get one, get this one. (And if you manage a DW, I'd seriously consider investing in a copy for every person on the project. I think you'll be glad you did.)

Calling this book "Object-Oriented..." is quite a stretch. This book has a several page description of object modeling. This description is the standard description that probably can be found in hundreds of other places. If you want to read an honest attempt to explain an object-oriented approach to data warehousing, read "The Data Warehouse Method".

Beyond that, the author has almost nothing new or particularly insightful to say about data warehousing. A large part of the book is given over to a discussion of dimensional modeling. Rather than build on what Ralph Kimball has written on the subject, the author just regurgitates what others have already written. If you want to read about dimensional modeling, read "The Data Warehouse Toolkit" and "Data Warehouse Design Solutions" rather than the retread material in this publication.

Great book. The author has provided a very good perspective on business intelligence and the issues relating to multidimensional data structures. Unlike others I've read on the subject, this book describes the actual process in detail. It gave me a great understanding of how to a set up a data structure based on the needs of the end user. I would recommend this book to anyone working in the world of business intelligence.

What this book offers is perspectives on designing warehouses with the goal of "business intelligence". Particularly useful is an interview process of extracting requirements from the end users, business strategists, upper management, and IT support people. Giovinazzo, also introduces the topology of multidimensional databases (e.g cubes, stars, and snowflakes) and the use of "fact" tables to tie these dimensions together. Through the use of the "Big Brother and the Building Company" example, he provides some down to earth examples. In a few appendices he introduces spatially enabled warehousing (weakly), metadata standards and good tips and tips on extraction, transformation, and loading. These alone are not "minor subjects", but warrant the introduction given here (and subsequent books). The books title includes "object oriented" but is perhaps not emphasized enough. I think the real value is getting the designer to think about the needs and dimensions early in the process.

Mr. Giovinazzo, titling this book Object Oriented Data Warehouse Design is quit a stretch. True it does contain the Object Oriented buzzwords, but no real practical OO design or implementation information. A better title would have been 'Introduction to Data Warehousing' but then this would just be remake of a hundred other books already available on that subject.

This is the one book that will explain to you what data warehouse design is all about. Many books have been written on relational database design. Many other books have been written on data mining. And so many others on data warehouse design. Yet data warehouse has been so little understood. It may be because there are too many good books on relational databases and not so many good ones in data warehouse. With this one book the scenario may change. Finally people will not have any more excuse to say that they do not understand what a good data warehouse should be like. And, most importantly, how much it differs from the well-known relational databases.

This book was very help for explaining the importance of data warehouse by functional users for business intelligence purposes. The book is laid-out in a systematic fashion that allows a non-technical person the ability to understand the process of building a data warehouse and the "technical" aspects of one. While technical terms are used through the book, the terms are defined in non-technical terms with at useful glossary at the end of each chapter for any new terms introduced in the chapter. The use of a detailed case study provides that reader with the ability to tie the "theory" with the "application" of a data warehouse. Also, the graphs and charts add in the explanation of many key ideas and issues with a data warehouse.

The only shortcoming of the book is that it does not address in any great detail how applications such as activity-based costing or balanced scorecard can use the data for analytical analysis.

Dear Mr. Giovinazzo, Your book as a primer is OK, but you could have done better! We skip the firs quick chapters on project management and the last ones on various minor subjects. The beginner has plenty of it in the ever growing Inmonian sub-literature. What a beginner needs is a good well cut problem ad its solution. You come up with two nice examples and work on them, discussing the main aspects of dwh design. From pages 92 to 232 you address some interesting problems and suggest the relative solutions. Unfortunately there seem to be some editing mistake here and there. For example text and picture don't match at page 182 and I would't recommend using dimension type 2 at page 164. It's just type 5 without user code and it's useless. Keep up the good work.

Roberto Botto, Rome - Italy

This book is a compliment to Kimball's book. The point of the book is not to provide astounding new star schema design revelations, the very point of the star schema is simplicity, but to provide a method of defining a star schema that meets the needs of the user. I agree with the Chicago reader that says that another good book is Kimball's. Both provide valuable insight. Kimball lays the foundation of the star schema itself, while this book demonstrates how to use object oriented methods to link that schema to the needs of the user.