Pado Plausibility Data Set

The verb-argument plausibility ratings used in Pado, Keller and Crocker (2006) (and later work) can be found here (PropBank and FrameNet 1.2 role annotations).

Annotations for McRae's Data

I'm happy to provide role annotations for the McRae data set upon request - please contact Ken McRae first for permission to use his data.

McRae Test and Development Split

McRae's data set was split into a development and test set in Pado, Keller and Crocker (2006) and later work. The development set (15 verbs x 4 argument ratings = 60 data points) consists of the verbs recognise, search, arrest, interrogate, investigate, hire, interview, study, carry, grade, examine, instruct, dismiss, sentence, question .