The Declassification Engine2013-2014

A partnership between faculty and students in the Departments of History, Statistics and Computer Science at Columbia University, this project will probe the limits of official secrecy by applying natural language processing software to archives of declassified documents to examine whether it is possible to predict the contents of redacted text, attribute authorship to anonymous documents and model the geographic and temporal patterns of diplomatic communications.  (Half of this project’s budget will come from the Tow Center.)

The Team