Corpora, evaluation material, annotated records and datasets used to develop and assess models.
Publicly accessible historical web crawl archives containing WARC and related data, with periodic releases that support web-scale text analysis.