2011
01.17

Natural Language Processing with Python is an extended tutorial using the Natural Language Toolkit (NLTK) Python library to explore Natural Language Processing (NLP) concepts.  It’s probably best suited for readers who already have a background in NLP but who want to learn Python/NLTK;  approaching the text as an experienced programmer curious about NLP, I found it disappointing.

The authors start out well, quickly establishing a working environment and providing  code examples using the NLTK library; note that you’ll need Python 2.x as NLTK is not yet ported to Python 3.  The library provides extensive test data and the exercises can be completed without errors.

Very early on, though, I found myself asking “why am I doing this?” as I completed a code sample.  As an example, it’s very nice that the NLTK library can display a dispersion plot, but what does this really tell me about the data, and more importantly, why do I need to know that?  Assuming that my lack of NLP background was the problem, I continued on, only to have the text jump to a discussion of Python functions and lists.   By the third chapter, I had lost interest.

I plan to review some of the background materials suggested by the authors in the “Further Reading” sections and possibly return to this book if time permits.  For now it remains mostly unread, as the alternating NLP/Python discussions just weren’t helpful to fully grasp either topic.

 

No Comment.

Add Your Comment