no code implementations • LILT 2016 • Lori Moon, Patricija Kirvaitis, Noreen Madden
The results show that human annotators have fairly good agreement on some of the categories, such as whether or not a use is epistemic, but poor agreement on others.