Data Mining with Microsoft® SQL Server™ 2000 Technical Reference (IT Professional)

By Claude Seidman

The amount of information stored in corporate databases is exploding exponentially. Data mining--finding meaningful patterns in all that data--can give any organization a competitive advantage. This... This description may be from another edition of this product.

• Format: Hardcover
• Language: English
• ISBN: 0735612714
• ISBN13: 9780735612716
• Release Date: May 2001
• Publisher: Microsoft Press
• Length: 384 Pages
• Weight: 2.10 lbs.
• Dimensions: 9.3" x 0.8" x 7.4"

Customer Reviews

Book is better than the product

I have to agree with one of the previous reviewers when he said that given the absence of practically *ANY* documentation provided by Microsoft, this book is your only real source of information about Microsoft's data mining product.I'm a big fan of OLAP amd data mining which made me better appreciate the time the author took to lay the groundwork for the discipline of data mining. Unlike a previous reviewer, I think that the author shares lots of real-world evperience which you can see by the way he bring up problems (which I have encountered myself) that occur when moving from raw data to a data mining model. He also catches some glitches and unreported features in the product for you and shows you how to work around them. The book is actually very complete considering that the data mining product put out by Microsoft is promising, but extremely rudimentary. It provides only two basic data mining algorithms and gives a very clumsy way to try to add other algorithms. Thankfully, the author discusses techniques and pitfalls of mining numerical data and even shows you how to use SQL Server 2000 to perform a regression analysis for that purpose.I would have given this book five stars except for two points :1: The mushroom database is a good illustration of the use of the decision tree algorithm, but I think it may have been good to include a more business-oriented example that would bring data mining closer to it's intended purpose. 2: I was a little disappointed not to see any explanation as to how to add your own algorithms to the data mining product. Even if doing so requires C++ experience, it would have been perfectly fine to include it in a separate chapter or in an appendix. I don't know why the author chose not to include it.Byond that, I would definitely recommend this book if you need to use MS data mining. The book is well written, and considering the infancy of the product, it's also very complete. Besides, you have no other real resource out there!

Working in a Data Mine

I always look for Seidman's contributions in SQL magazines and at conferences since he always has something interesting and innovative to say that is fuelled by his depth of knowledge of the subject. What's more he has a great knack of presenting complicated (and sometimes, let's face it, not very exciting) concepts in a way that makes them easy to understand and in contexts that anyone can relate to. In particular his style is reminiscient of Roger Sessions (Com+ and the Battle for the Middle Tier, etc) minus the cynicism but with equal enthusiasm. No book can be all things to all people, but this is an excellent introduction to the world of Data mining and the power behind SQL200's implementation of it. I would recommend it to anyone looking to discover those hidden trends and patterns in their data, exploit them and become their CFO's best friend.

Extremely useful

I found this book to the most helpful resource on MS data mining out there today. I was afraid I'd find a rehash of the Microsoft help files, but was pleasantly surprised to find lots of new, original information. There's lot's of dicussion of process as well as step-by-step instructions on the use of the data mining product. Very good explanations as to the use of DSO and DTS to create models and to query them.