Attend the Women in ML Symposium on December 7 Register now

wsc273

  • Description:

WSC273 is a common sense reasoning benchmark that requires the system to read a sentence with an ambiguous pronoun and select the referent of that pronoun from two choices. It contains the first 273 examples from the Winograd Schema Challenge. A Winograd schema is a pair of sentences that differ in only one or two words and that contain an ambiguity that is resolved in opposite ways in the two sentences and requires the use of world knowledge and reasoning for its resolution. The schema takes its name from a well-known example by Terry Winograd: The city councilmen refused the demonstrators a permit because they [feared/advocated] violence.'' If the word isfeared'', then they'' presumably refers to the city council; if it isadvocated'' then ``they'' presumably refers to the demonstrators.

Split Examples
'test' 273
  • Feature structure:
FeaturesDict({
    'idx': int32,
    'label': int32,
    'option1': Text(shape=(), dtype=object),
    'option1_normalized': Text(shape=(), dtype=object),
    'option2': Text(shape=(), dtype=object),
    'option2_normalized': Text(shape=(), dtype=object),
    'pronoun_end': int32,
    'pronoun_start': int32,
    'pronoun_text': Text(shape=(), dtype=object),
    'text': Text(shape=(), dtype=object),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
idx Tensor int32
label Tensor int32
option1 Text object
option1_normalized Text object
option2 Text object
option2_normalized Text object
pronoun_end Tensor int32
pronoun_start Tensor int32
pronoun_text Text object
text Text object
  • Citation:
@inproceedings{levesque2012winograd,
  title={The winograd schema challenge},
  author={Levesque, Hector and Davis, Ernest and Morgenstern, Leora},
  booktitle={Thirteenth International Conference on the Principles of Knowledge Representation and Reasoning},
  year={2012},
  organization={Citeseer}
}