28 of 31 people found the following review helpful:
3.0 out of 5 stars
Useful Text, Marred By Poor Technical Editing and Typos, December 2, 2005
This review is from: Data Mining with SQL Server 2005 (Paperback)
This could have been a 5-star book were it not for the numerous editing oversights. Note that I say "editing oversights" - I don't really blame the authors as they have the right credentials for tackling this subject.
Examples of editorial omissions:
1) Page 45 - the text says that the MemberCard_Prediction mining model uses Gender, Age, Profession, HasChildren and HouseOwner to predict the membership card type. However, in the definition of mining model itself, HasChildren attribute is missing.
2) Page 46 - same mistake as #1
3) Page 50 - poor editing - "the result rowsets has the structure displayed in Figure 2.4". Great. Now where exactly is Figure 2.4? You have to flip pages in the book back to page 41 to find it. How much effort would it have taken for the editors to rephrase the sentence to "the result rowsets has the structure displayed in Figure 2.4 on page 41."?
3) Page 51 - not really a mistake, but poor editing anyway - "In Figure 2.7, the table on the right is a truth table. The left table is a new......". Now, when you look at figure 2.7 on page 52 and see the way tables are arranged, it would have been better to say "The bottom table is a truth table and the top table is a new....).
4) Page 53 - "A Prediction Query Example" - the Select statement refers to M.MemberCard, but "M" itself not listed as an alias in the From clause.
Ok...I think you get the idea. Now to the nice stuff:
It you are interested in data mining with SQL Server 2005, this is still a book you must have. Those with an understanding of data mining principles will benefit most. In addition, you may need to brush up on statistics to really understand what is going on.
Chapter 1 - Introduction to Data Mining -quick intro to data mining, major vendors, project cycle
Chapter 2 - OLE DB for Data Mining - good coverage (despite the errors) of key concepts. You may be tempted to skip this chapter and dive into the "newer" stuff in Chapter 3 - I would urge you to understand chapter 2 before you move on.
Chapter 3 - Shows how to use the Business Intelligence Development Studio
Chapters 4 - 10 - Cover the various DM algorithms from Naïve Bayes to Neural Network. Includes both the general description of the algorithm as well as syntactical stuff as applied to SQL Server 2005.
Chapter 11 - Mining OLAP Cubes - fairly rudimentary stuff here.
Chapter 12 - Data Mining with SQL Server Integration Services (SSIS) - introduces SSIS and then covers some basic ground relating to DM tasks in SSIS.
Chapters 13 and 14 - SQL Server Data Mining Architecture and Programming
Chapter 15 - shows how to implement a simple web cross-selling application. Don't expect to implement Amazon like recommendations, but it is a good start anyway.
Chapter 16 - discusses using Excel for forecasting using Analysis services in the background
Chapter 17 - A brief (about 10 pages) text devoted to extending the DM framework using custom plug-in algorithms.
Help other customers find the most helpful reviews
Was this review helpful to you? Yes
No
19 of 21 people found the following review helpful:
3.0 out of 5 stars
Decent Book, March 16, 2006
This review is from: Data Mining with SQL Server 2005 (Paperback)
At the time I'm writing this review this book is the only one available dedicated solely to the data mining features of SQL Server 2005. The book is good, but I was disappointed in it as well on three fronts. First, there is a chapter dedicated to each of the data mining algorithms. I didn't find the business use case examples for when, why, and how to use each algorithm sufficient. Second, each of the algorithm chapters goes off the deep-end explaining the mathematical basis for the algorithm. There are very, very few developers who are going to remember enough of their college mathematics to follow along. Third, the technical coverage of how to use the APIs and the data mining extension language (DMX) is superficial, particularly with DMX. After reading this book cover to cover I couldn't go off and write a DMX query if I wanted to. On the application I'm working on we are planning to implement our own web visualization viewers for the data mining algorithms. This book didn't give me what I needed in understanding the object model exposed by the APIs in order to handle the back-end coding for this. All in all, if you are planning to do data mining with SQL Server 2005 I would recommend this book only because at the time of this writing there is nothing else available. However, you will learn quite a bit about data mining with it and depending on your prior experience (more is better) it might be an excellent fit for you.
Help other customers find the most helpful reviews
Was this review helpful to you? Yes
No
8 of 8 people found the following review helpful:
4.0 out of 5 stars
Good book, terrible editing, January 23, 2006
This review is from: Data Mining with SQL Server 2005 (Paperback)
Another reviewer mentioned poor technical editing. My impression was that the code samples perhaps had a few syntax errors and wouldn't run without a bit of editing, and that the reviewer was probably being overpicky.
My impression was wrong! This is the worst edited book I've ever read. There are mistakes throughout that seriously distract the reader from the content.
This is unfortunate, because I thought the content was very good. If you are a statistically training data analyst, it might not go into enough depth, but if you are a database developer looking to bring data mining to your business/applications, this provides excellent coverage, from how to create the models to good descriptions and comparisons of the different algorithms SQL 2005 offers.
Help other customers find the most helpful reviews
Was this review helpful to you? Yes
No
