I am trying to get statistics on 2-character sequence combination frequencies for the English language for a project I'm working on. It's very important that the results are as precise as possible. I'm not quite sure whether this type of research has been conducted before, but not to worry; I can easily whip up a script to report the statistics I need.
But what I don't have is a large body of sample text to analyse. Does anyone know of a good resource where I can get a large amount (several GB would be nice) of English text covering a wide variety of topics in plain-text?