Graph-Convolutional Autoencoder Ensembles for the Humanities, Illustrated with a Study of the American Slave Trade

We introduce a graph-aware autoencoder ensemble framework, with associated formalisms and tooling, designed to facilitate deep learning for scholarship in the humanities.