Trying to figure out which combination of corpora datasets will get me the best procedurally generated fake company names.

From: https://github.com/dariusk/corpora/tree/master/data