America's Public Bible is a project to detect the biblical quotations in the Chronicling America and 19th Century U.S. Newspapers datasets of historical newspapers, then to interpret and visualize the patterns.

This repository contains the code that extracts the features, trains the models, and finds the quotations.

• The bin directory contains R scripts that are intended to be run on newspaper batches, as well as the shell scripts to run them on the HPC cluster.
• The model directory contains code and data to train the prediction model.

This repository has undergone a number of changes. The initial code for the prototype version of the site was created in 2016. That code can be found in this tag on the repository. Much of that code has been superseded, and what remains in the master branch is located in the prototype/ directory.