I'm reading this now. Things like this remind me of why I love the Internet. So much great knowledge, freely available for everyone. It boggles the imagination sometimes.

Also, another book that some HN'ers might find interesting is this one:

Information Theory, Inference, and Learning Algorithms:

http://www.inference.phy.cam.ac.uk/itprnn/book.html

Another one:

The Elements of Statistical Learning: Data Mining, Inference, and Prediction.

http://www-stat.stanford.edu/~tibs/ElemStatLearn/

I'm trying to collect a list of resources of this quality in this subreddit: http://www.reddit.com/r/learnit/

Related course: http://cs276.stanford.edu and syllabus: http://www.stanford.edu/class/cs276/cs276-2009-syllabus.html

This is a fantastic book, I'm using it to build my own IR system and it has been immensely useful. I also own Managing Gigabytes but I find this book to be much more useful and up to date.

Another +1 on this fantastic book. It's extremely accessible and concise while covering quite a lot of ground on the subject.

This is a fantastic survey book. It's one of those few books that's exactly what it's called. You can always dive deeper on the stuff that's interesting/relevant to you after reading this book, but I have yet to find a book at this level that's this good.