The complete corpus: 103.1 billion tokens, 408 million posts, 18,347 newsgroups, 1980–2013.
This is the gated full release of the dataset sampled at
OwnedByDanes/Usenet-Corpus-1980-2013.
Evaluating? Don't request access here — use the sample. The sample files are
free, ungated, and use the identical schema and cleaning pipeline. This repository
exists to deliver the corpus to licensees, not to be browsed.
