A corpus is a large collection of written, spoken, or otherwise recorded language material used for analysis or model training. Corpora support tasks such as language modeling, translation, information extraction, and linguistic research.