Although my postings refer to information design for print, there is a lot of interesting work in online (information/ data) visualisation.
You can see a bunch of them here at Information Aesthetics.
This basically involves the designer/ engineer creating an ongoing design system to show patterns and hierarchies. It needs flexibility above and beyond what can be frozen on a printed page and all this while dealing with the need for the user to interact with this data too, after an initial overview. (I personally think an initial orientation is important - and then the ability to drill down - others do not and seem to revel in initial complexity.)
One method used to show links associated with a website is the tag cloud. And there is a new website called Wordle that lets you make similar typographic arrangements based on the text from any given source.
These are actually visual displays of words that appear often in a body of text and so an indicator of what is important to the author. I put the contents of the current wikipedia WWII page through it and the results are here.
My first thoughts are that it is good to see the word Soviet so prominent, the USSR often being disproportionately represented by much western media on the war. Japan is surprisingly prominent - I wonder if that indicates a US-leaning view of the war (much of the US involvement was the massive Pacific-wide naval and amphibious war with Japan). And interesting to see China so much more prominent than in the books I grew up with.
Wordle is not an analytical tool and to be so would need to stress sentence and grammar intention and structure and context etc etc but it is a lot of fun and will do a serious job of bringing these concepts to a larger audience. It is easy to use and if you want to find out about the creator, here is his blog.
I should mention the work of Stefanie Posavec if you want to look at further visual patterns within text.
And in case anyone is interested, this site's Wordle cloud is like this.