Led by Matthew Connelly at Columbia University, the Declassification Engine project is a wide-ranging effort involving "historians, journalists, legal scholars, statisticians, and computer scientists joining together to [explore] novel statistical/machine learning methodologies" focusing on "attributing authorship to anonymous documents, characterizing attributes of redacted text, and modeling spatial and temporal patterns of diplomatic communications." Already the project has demonstrated using big data approaches to unredact documents that have been released piecemeal over long periods of time, to predict the contents of fully redacted passages, and to visualize global patterns of American diplomacy. Kalev helped create the Sphere of Influence visualization. The project has been extensively covered from Wired to CJR, to The New Yorker.