Customer Reviews

Data Mining: A Tutorial Based Primer

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

 

 

The book does a good job covering basic data mining terminology and concepts, and provides introductory, but still subtantial coverage of the most common data mining methods.

Positives: a trial version of the easy-to-use Excel-based iDA tool is included with the book, which allows the reader to reproduce the examples (very helpful for understanding the text). iDA may also be used to complete many of the well thought out exercises provided at the end of each chapter. Working with the hands-on examples and exercises is an excellent way to learn data mining, and due to this, the book provides unique and excellent value.

Negatives: the order of topics and chapters seems rather disorganized; topics are often (surprisingly) repeated and the overall structure of the book doesn't seem to make sense at times. But if you read each chapter more or less independently, this isn't a serious problem. The iDA tool that comes with the text is a trial, 180-day version, and it is unlikely that the average reader will want to spend $5,000 to purchase a license for the commercial product after the six month trial is up. So you should buy the book knowing ahead of time that after a while, the iDA tool will no longer be available to go back over the examples or exercises.

The particularity of this book is that it is more accessible to read than most of data mining books, which in general require some maths/statistics/computing background.

The book is not written in the best way from the point of view of a data mining expert, as for instance sometimes a theme is recurrent in the text, but it is not obvious to explain data mining concepts using minimal previous knowledge in computing/maths/statistics.

A second important positive aspect is that the book comes with a software (IDA) running under Excel, which can be used to illustrate the techniques presented in the book (BTW a new version of the software is freely available to download, regularly). This is not the case with most of the data mining books. So if you wish to learn the basics of data mining with minimal or no previous resources (good maths/computing background and access to expensive data mining software) then this is a very good choice.

Using an Excel-based tool to provide a hands-on approach, this book covers diverse areas in data mining from statistical patterns to data warehousing to the semantic web. Easy to read and with many insights.

When I bought this book, I knew nothing about data mining. Unfortunately, this book glossed over the topics I knew least about and spent a depressing amount of time on stuff anyone should have learned by junior high. They introduced 188 "Key Terms" in a book that's only 350 pages long. In chapter 1 they give definitions for words like "fact", "hypothesis", etc. Yet by chapter 5 they start flinging the symbols for attribute standard deviation at you with no explanation or warning. So I'm not sure who they think will be reading their book...but from what I can tell, they assume their target audience can handle advanced algebra with ease but may need a definition of "the scientific method".

They also spend quite a bit of time walking you through the Excel PivotTable creation wizard and other such fluff. They carefully instruct the reader that dragging and dropping is accomplished through use of the mouse, and that you should drop columns into the area marked 'Drop Column Fields Here'.

On the upside, I do know a bit more about data mining now. I don't feel that I could run right out and get a job, but at least when I start reading another book I'll have an idea of what the terms and concepts are.

So I suppose if you're good at statistics, have never taken a basic science course and have poor computer skills, this book is for you.

The iData Analyzer Software from infoacumen.com on CD included with the text is problematic at best. Students in my Business Intelligence and Data Warehousing courses have had many problems with installation and with running Data Mining Sessions even with the included sample worksheets. Downloading the 'latest' version of the software directly from the infoacumen web site did not improve the situation. Inquiries to the infoacumen site support email address have gone without reply. Software runs in 'compatibility mode' in Excel 2007 and even with macro security turned off, Excel often crashes. Data Mining Sessions often fail with a variety of error messages even with running the supplied sample worksheets. The Excel add-in interface is completely devoid of any GUI and no graphic output of Data Mining Sessions are available. To work around these problems I have had to ferret out sample worksheeets that would not crash Excel and post the Data Mining Session results for review and interpretation. As an alternative we have made extensive use of Microsoft SQL Server 2008 Analysis Services Data Mining Project Options and Tools and Oracle Data Miner software.

The text offers a fairly good treatment of Data Mining, however many topics are presented in a manner and discussion over complicating various subjects.

The author is not able to clearly present/describe concepts without using many examples. For many subjects(even simple ones), the book is lacking in clarity and logic, so that it's hard to find out what the author is trying to say. Too many pages are wasted on irrelevant sentences. Don Box books contain difficult logic, but you feel his clarity and preciseness in his writing. George Shepherd books are also enjoyable to read even though the subjects covered are wide and not easy. This book is just the opposite: easy concepts presented with a messy, hard-to-follow style.

This book and CD was probably a good purchase "back in the day", but I am taking a DW class now (2011) and I am here to tell you that if you have any operating system newer than Windows XP, the IDataAnalyzer is not going to work, at least not for me, I even tried installing under XP Mode and I was not successful on my Windows 7 Machine. Our class had to install Weka, which is good, but we can't successfully follow the examples in this book due to too many differences. IDataAnalyzer is an Add-In to Excel and so far I can't get WEKA to work that way, so in a nutshell, my school made a bad decision to have us purchase this book due to the fact most people have OSs newer than Windows XP. I think the book is great, but only for people who have XP. Oh, there is too many darn formulas in this book. I am still not sure why, but I am new to data mining so maybe I am missing the point of all the formulas.

This book is touted as an intro to Data Mining, and while it does cover most of the intro topics you'd want to learn about, it seems to cover some in more depth. This was a text book for an Information Organization/Retrieval course I took, but our instructor subsidized it pretty heavily with other sources and her own notes. Worked very well when used in a classroom setting.