I second this :) if you're interested in more about n-grams you might wanna read Manning S. , Schutze - "Foundations of statistical natural language processing" chapter 6