Authors: Steven Bird, Ewan Klein, and Edward Loper
Format: Paperback, 502 pages
Publisher: O'Reilly Media; 1st edition (June 30, 2009)
ISBN-10: 0596516495
ISBN-13: 978-0596516499
According to Wikipedia, "Natural language processing (NLP) is a field of computer science and linguistics concerned with the interactions between computers and human (natural) languages. Natural language generation systems convert information from computer databases into readable human language. Natural language understanding systems convert samples of human language into more formal representations such as parse trees or first-order logic structures that are easier for computer programs to manipulate. Many problems within NLP apply to both generation and understanding; for example, a computer must be able to model morphology (the structure of words) in order to understand an English sentence, and a model of morphology is also needed for producing a grammatically correct English sentence". OK, did you get all that?
According to the blurb on the book's back cover, "This book offers a highly accessible introduction to natural language processing, the field that supports a variety of language technologies, from predictive text and email filtering to automatic summarization and translation". Putting it all together, this book teaches the reader how to do something highly specific and practical with the Python programming language. I put it this way because there are many books out there that will teach you how to learn Python, or some other programming language, which includes learning various typical tasks, but they don't teach you where to go next.
You've probably read various "adverts" about Python saying that YouTube, George Lucas's Industry Light and Magic, and the SSL certificate company Thawte, among others, have built their success on Python, attempting to induce developers to learn and use Python (assuming they don't do so already). Here we have a book that says something like, "This is what we use Python for and you can learn to do the same thing with Python", again, assuming that you need to do such a thing.
If you read the entire Wikipedia article about NLP at the link posted earlier in this review, you'll see that this is a very dynamic field with many applications, so writing a book on the topic and supporting the use of Python for NLP is a very reasonable thing to do. If you're involved in NLP in any manner or fashion, you might want to have a look. The Preface states that the book serves a wide range of people who need to "learn to write programs that analyze written language regardless of previous programming experience".
The early chapters of the book are written for the programming newbie, so even previous programming experience is not absolutely required. Once you work past those chapters, Python as a specific language is introduced, so any coders interested don't need to know Python already to pick up this book. All that said, it looks like being a quick learner would be extremely helpful, especially since your primary goal isn't to learn programming or even to learn Python, but to use Python as a tool. That's sort of like saying you don't need to know how to use a hammer to hit a nail in order to buy a book about building houses. It's possible, but something of a stretch.
I think you'll get the most out of this book if you have at least some programming, and preferably some Python experience, and treat this book as an introduction to NLP. It doesn't really make much sense otherwise. I wouldn't try to use this book to learn Python if I didn't have an overriding interest in NLP. In fact, if you don't have much in the way of Programming/Python experience, you might want to consider picking up a few other books first, learn the foundation of programming and Python, then come back and start where you left off with the NLP book.
The book's authors didn't lie when they said this book is meant to be a practical introduction to NLP. It's long on code, but a little short on theory. I've gone through plenty of programming books that were the opposite and they nearly bored me to tears. Not that theory isn't useful by any means, but programming really needs to be performed in order to be used.
I mentioned learning programming and Python from other sources first, because a good grounding in both the theory and practice of said-topics will make it a lot easier to apply them to NLP. Although I'm not expert in the field, it might also be better to seek out some more generic NLP texts to tackle the theory behind the discipline before hitting this book. If you're very smart and you're very quick, this book may suffice, but if you're otherwise (I don't mean "dumb", just not brilliant), creating a solid base on which to stand Bird's, Klein's, and Loper's material might work out better for you.
The book is meant to stand alone, but as I said earlier, this is a specific field and discipline that can be applied hundreds of different ways. If you're still unsure about buying this book after reading the review, look it up online, thumb through it at your local bookstore or, heaven forbid, read a few more reviews to get additional perspective, then make up your mind. Unless I miss my guess though, if you are involved in NLP at all, sooner or later, you'll need this book in your library.
