I’ve subscribed to the RSS feed(s) here and have seen regular posts of NLP book titles, but they aren’t accompanied by a rating or recommendation. Most of these $100+ books are written in the impenetrable prose of a doctoral thesis. Math and industry jargon are fine, but I can’t stand sentences like this:

The principal purpose of this monograph is to open for Systems Science a new field of studies – the development of formal models and methods intended for helping the designers of semantics-oriented NLPSs to overcome numerous problems of logical character associated with the engineering of such systems.

Four lines of blather to tell me that the book introduces new concepts. A poorly written book on the topic of processing natural language kind of erodes the credibility of the author, don’t you think?

There must be clearly written books on NLP. I’d love some recommendations.

This is where the good stuff is at.

http://www.cs.rochester.edu/~james/

James Allen’s book “Natural Language Understanding” is still the best book on the subject by a considerable margin.

Although the second edition is more than a decade old now, it is still as good as anything current because most researchers have been concentrating on statistical natural language processing in the intervening time, which is actually a dead end.

As a general observation I’d like to add that while some doctoral thesis prose is not as well written as it could be, you’re still going to have to come to grips with the mathematical and symbolic languages commonly used to describe algorithms, theories and supporting proofs. However you shouldn’t have any trouble with the aforementioned book—it is beautifully and clearly written, and contains an immense amount of immediately useful detail, as well as comprehensive summaries and introductory material.

Thanks for the recommendations! Mathematical/symbolic language is definitely not a problem; it’s the unnecessary verbosity that I can’t stand.

I found it interesting that Allen’s list of publications seems getting smaller and smaller each year. Has the field of NLP cooled off for some reason?
http://www.cs.rochester.edu/research/cisd/pubs/

I was happy to see “Natural Language Understanding” available used for US$4.00 on Amazon.
http://www.amazon.com/Natural-Language-Understanding-James-Allen/dp/0805303340/ref=sr_1_1?ie=UTF8&qid=1317040775&sr=8-1

I also picked up the “Handbook on Natural Language Processing, 2nd Ed”, which appears to be a compilation of current thought as of 2010.
http://www.amazon.com/Handbook-Language-Processing-Learning-Recognition/dp/1420085921/ref=sr_1_2?s=books&ie=UTF8&qid=1317040808&sr=1-2

Thanks again for the help!

Hi Brien,

Thanks for your feedback!I do fully agree with you! So we’ll definitely take your request into account. We have a few priorities however:

-first launch our commercial program to generate some revenues
-new home page
-serious forum upgrades
-socializing the website. The thing is that we different audiences on this website: scientists, commercial AI programmers, amateurs and in the future animations, speech & biometrics expert as well.

What we’ll need to do is make a rating per subcommunity, or even better: link to Facebook Friends/LinkedIn connections. We can then make recommendations like: Your Friend x,y and z liked this book. Or: people in your conversations (like this thread) liked this book.

It will take a while though, but it’s certainly on our wish list!!

If you have other feature request, please leave them in our feature request forum:
http://www.chatbots.org/admin/viewforum/13/

Hi Erwin—I didn’t realize I was making a feature request! LOL Thanks for adding ratings to the to-do list!