Words Vote.
- App URL
- http://code.google.com/p/wordsvote/
- Source URL
- http://code.google.com/p/wordsvote/source/browse/trunk/
- Description
"Words Vote." is a client-side python program that employs Sunlight Labs' "python-capitolwords" library to obtain lists of words used by individual congresspeople during an interval of time (usually before a major vote). Also extracting data about a selected roll call vote from Govtrack's XML records, the program uses Bayesian statistics to assess which words are most informative in predicting a congressperson's vote.
Although linguistics researchers already have access to established text corpora, the growth in standardized data formats and accessible APIs is permitting machine-friendly, real-time corpora to emerge and be used by non-professionals for political or casual research. By linking various syndicated categorization systems (here voting records and real-time text corpora), machines can help us identify many interesting and hidden associations in political speech.
And the better we understand "mere" political rhetoric, the faster we can move beyond it.
This application was submitted by Todd Fine on March 31, 2009