The use of rating and Likert scales in Natural Language Generation human evaluation tasks: A review and some recommendations.

Jacopo Amidei, Paul Piwek, Alistair Willis

Browse the full INLG paper archive.