Customer Reviews

Data Mining: Concepts and Techniques, Second Edition (The Morgan Kaufmann Series in Data Management Systems)

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

 

 

This is a great textbook for an undergraduate or layperson to the information sciences, but specialists may find it lacking depth. It is very good at identifying practices and principles that would guide a high-level planner toward a sound research program. That said, this book exhaustively covers the breadth of the modern field at the expense of formulas, algorithms, and source code that would have been valuable to an engineer or scientist with plans to implement.

* Buy this book if you require a high-level understanding of the concepts and techniques used in the field.

* Don't buy this book if you are planning to specialize in data mining, or if you have plans to implement yourself.

I have read the first edition of this book years before. This second edition has significant improvements. Core topic (classification, clustering, association rules) is very detailed and much easier to read. The author also add much material about advanced topics such as graph mining, multimedia mining, stream and time series mining, etc. Although these advanced topics are not as well writen as core topics, at least you will get idea about what's going on in these areas.

This is a useful book: it provides the most comprehensive state of the art overview of data-mining technology I know of. The emphasis is on 'overview' however - you can find starting points and intuitions, but you will not be able to to do anything very ambitious just on the basis of the purely technical information here. At one point, the details of how linear classifiers work are swept under the carpet with a faintly crass remark about 'fancy math tricks'. If linear classifiers are 'fancy math tricks', what does that make variational methods for probabilistic data modelling? Note, in fact, that advanced machine learning in general, where fancy math tricks are ubiquitous and unavoidable, is not touched - an interesting implicit distinction.

Further, this is not a book you are likely to read for pleasure, for either the prose or the presentation. If you are not professionally involved, you neither need nor want it.

Nevertheless, given all those reservations, I'm happy to have it on the shelf.

Dr. Han is a leader in Data Mining; but unfortunately this book does not speak for that. The explanation is poor, the language is weak and thus, the book is not at all a good read. The book by Pang-Ning Tan and Kumar is much better.

The only good thing is that the second edition has a comprehensive coverage and contains many recent topics (streaming, social network, etc.) which is not available in other textbooks.

I bought this book as a text book for data mining. I found this book give a solid introduction to multiple topics and a ready reference. One thing , I found though was a rather superficial treatment of very specific algorithms and a thorough treatment of general ones . Atleast the most popular specific algorithms can be detailed.

If you have a professor that assigns this book for your class, run the other way. This book is extremely frustrating. It only explains in detail the easier concepts with examples,(which are incoherent), and neglects explanation and examples on all the advanced topics. It is full of formula that is incomplete, along with confusing notation.

As wrote in title, it is really good book and fast shipping. I am satisfied!

Thank you!