Customer Reviews

Data Mining with SQL Server 2005

5 star:		(8)
4 star:		(2)
3 star:		(3)
2 star:		(0)
1 star:		(3)

 

 

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.

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.

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.

Some people commented on the poor editing: typos & some wrong pictures. True. (It detracts, but you can figure it out easily)
Some stated that it is not a good general overview of Data Mining. True (though it has a bit of a summary)
Some stated that is doesn't discuss business applications in detail. Yes, (it only makes brief reference to them).
Some stated that it is very vendor specific. Hello, read the title - SQL 2005.

It is a must read for anyone who wants to take maximum advantage of SQL Server 2005 Data Mining. It goes thru all the algorithms, tells you how each one works, how to tune them & how to embed them into your applications. It compliments the Books On-Line materials, tutorials & sample code that ship with the product.
(interesting how people pay for a textbook & never bother to read the copous amount of materials that ships with the product.)

It does give you a bit of background in DM, & does walk you thru using the tools (SSMS & BIDS) used to create & administer the Data Mining.
It doesn't talk about using the Data Mining Viewer controls in Visual Studio 2005.

It is an easy read & very informative. Especially if you go to the trouble of downloading the samples & data from the web site & actually build the models with the book & step thru the code.

While it isn't really an indepth treatment of DMX in the way that "George Spofford's MDX Solutions" is for MDX. It does give you more than enough examples to be able to create, train & predict from the models.

It also gives enough to embed your DM models into your applications, Use them from Excel & take full advantage of the DM built-into SQL Intergration Services.

If you want an DM Overview for business use - check out Barry Lindof's book

I wouldn't normally review my own book, but I think the previous reviewer missed the point. This book is clearly about data mining in the context of the features of SQL Server 2005. It does not pretend to be anything else, and pretty much states so directly in the title. If you are looking for a generic, vendor neutral "all-up" book about data mining, its methods and applications, you're right, this book is not that book. However, if you want to learn about the data mining features of SQL Server 2005, how they work, how to use them and exploit them for your own analyses and applications, this is the book to read.

This is a great data mining book for database developers/DBAs, especially for those who are using SQL Server database.
There are simple and clear explanations of several popular data mining algorithms,plus lots of examples about how to solve business problems using these algorithms, code examples for SQL Server. This book is a must for those who wants to use SQL Server data mining for their projects. There are still some editing errors in the book.

There are precious few books written about Data Mining and this is the only one about SQL Server 2005 specifically. Beyond its originality, it is concise and well written. The first chapter is an introduction to Data Mining and it details exactly how Data Mining can be used in many different application scenarios. The remainder of the book is a detailed examination of all the new features and algorithms available in SQL Server 2005. It contains practical and relevant examples in an easy to read and intuitive style.

Even though the basic concepts involving Data Mining remained the same between SQL Server 2000 and 2005, there were significant enhancements made to the number and complexity of the data mining algorithms and the tools used to interpret data mining results. Without the inside knowledge provided in this book, even seasoned data mining users may miss some important opportunities. I would strongly encourage anyone even remotely interested in data mining to get this book and start discovering how data mining can separate their applications from all the others.

1) The book uses many examples based on the AllElectronics database. However, the book does not provide the AllElectronics database. Nowhere on the net where you can download AllElectronics database, even on the publisher website. This book is supposed to show us how to design and run data mining on SQL Server 2005, but it fails to do that since the examples in the book are not runnable. I've read many books about Data Mining for SQL 2005. All of them have runnable code except this book.

2) At page 45, the author used the Customer_id as a data column in Customers table in database schema of a mini data mart, but later on, in the examples, he uses CustomerId as data column. In many places in the book, I found many inconsistencies between the database schema and the column names used in the query. Since the authors did not bother to provide the schema of the database, it is very confusing to keep track of what is going on after an MDX query.

I have only recently started to get involved with Data Mining. I have been doing back end work with Analysis Services for a couple of years and we're ready to move on to the next level.

This book was amazing! The background in Analysis Services and Databases helped a lot, but the book covered all the topics in an easy to understand order. Sure, the chapters on the different algorithms can be very in depth, but apart from explaining the actual mathematical formulas, there is a huge amount of information about each algorithm that each developer MUST use when designing a Data Mining solution.

One of the reviewers commented that they can still not write a DMX statement. I'm confused by that statement! I am writing DMX statements using only the information I got out of this book. Sure, there are a few spelling mistakes here and there, but using the sections in the book where the syntax is fine, I've managed to run all the queries without issues.

I highly recommend this book.

I was really impressed with this book. It had a great introduction to demystify the typic of data mining. Since the learning curve on this topic is so high, these first few chapters are essential. It then immediately jumps into a practical example to help the reader bring it all together. The chapters get progressivily more difficult through the book and there's a chapter for each of the algorithms. The author team did a fantastic job and I'd highly recommend it.