In linguistics and natural language processing, a corpus or text corpus is a dataset, consisting of natively digital and older, digitalized, language resources, either annotated or unannotated.