Phrase similarity data
----------------------

This archive contains phrase similarity data reported in Gershman & Tenenbaum (2015). The data are contained in two files:

"phrases.csv" - each row corresponds to a single phrase
Column 1: phrase
Column 2: set (25 distinct phrase sets)
Column 3: type (1 = base sentence, 2 = meaning preservation, 3 = noun change, 4 = preposition change, 5 = adjective change)

"rankings.csv" - each row corresponds to a single phrase, ranked by a subject with respect to the base phrase of its set
Column 1: ranking
Column 2: set
Column 3: type
Column 4: subject number

REFERENCE: Gershman, S.J. & Tenenbaum, J.B. (2015). Phrase similarity in humans and machines. Proceedings of the 37th Annual Conference of the Cognitive Science Society.

Questions: please contact Sam Gershman (gershman@fas.harvard.edu)