Phrase Net
with Frank van Ham and Martin Wattenberg

Phrase Nets use a simple form of pattern matching to provide multiple views of the concepts contained a book, speech, or poem. The image below is a word graph made from Jane Austen's novel "Pride and Prejudice." The program has drawn a network of words, where two words are connected if they appear together in a phrase of the form "X and Y":



For instance, "Jane" and "Elizabeth" are connected by a thicker arrow since the phrase "Jane and Elizabeth" occurs 10 times in the novel. The result of this simple pattern matching scheme is a surprisingly coherent view of some of the concepts in the books. A large cluster of the main characters and their relationships is on the left; separate clusters touch on emotion and attitude. Smaller connected pairs ("fortune" and "consequence") touch on other themes.

Pointing another phrase filter to the same text yields a rather different view of Pride and Prejudice. Here's a phrasenet of "X at Y," which reveals some of the most important locations in the book.

 

Looking at different texts through the same phrase can also be quite illuminating. Below are two diagrams showing "X of Y" in the Old and New Testaments respectively.

For more on how the visualization works, please refer to this explanation page on Many Eyes.