In linguistics and natural language processing, a corpus (: corpora) or text corpus is a dataset, consisting of natively digital and older, digitalized, language resources, either annotated or unannotated. Annotated, they have been used in corpus linguistics for statistical hypothesis testing,