The internet and the web have become a wonderful field for applying mathematical methods to huge amounts of uncontrolled information. The ultimate contribution of this book, I think, is to present techniques by which results that are meaningful and useful for people -- such as text analysis and categorization, or relevance of web pages, or recommendation systems -- can emerge from the application of cold mathematical formulas. The mathematical tools are not necessarily new and may be hard to follow by non-experts (this is mostly Machine Learning, probability, etc). To fully appreciate the book, one needs to grasp the math, but I think the text is sufficiently interesting so that the book can also be enjoyed by people who don't like to read the math. The book is fairly comprehensive about modeling important processes that happen everyday on the web.