Keywords

This chapter tackles the task of keyword extraction from corpora. Keywords are extremely helpful to quickly identify the terms (and their associated concepts) that somehow define what a corpus is about. After a quick revision of the concept of keyword, I focus on the different methods that have been proposed to extract keywords effectively and efficiently. A key distinction is made between the reference-corpus method traditionally employed in corpus linguistics and the various methods that have been proposed in Natural Language Processing research. Through several experiments, the CCTC is explored using some of the most outstanding methods proposed to date, and a contrastive description of the results is offered.