These are a few of the tools and projects I've worked on recently. All are works in progress, and some are more exploratory or experimental than others. They can help analyze a text or set of documents, or just serve as fun tools to generate new ideas. Go ahead and try them out.
This maps the key words and the progression of the text (its "plot") in a two-dimensional space.
This automatically scales a set of documents and their words according to the major dimensions of variation, some of which are often ideological.
This characterizes a set of documents by finding a latent set of topics, along with the words associated with those topics.
This finds small clusters of related words (motifs) in the text.
The generates more text in the style of the original text -- although the syntax is a bit rough.
Choose Your Own
This is like Text Predictor, except you can choose which of the probably next words you want.
This counts "cliches," loosely speaking. It tracks the appearance of common 4-word sequences in your text.
This "summarizes" a text by finding the N most representative sentences.
This allows you compare a set of different texts, ranking their overall similarities in various ways.
This allows a group of participants to collectively write a text using a nomination and voting system that ensures strictly democratic equality. (Due to a change in PHP since this was written in 2002, existing projects may be viewed, but new ones cannot be created or modified.)