Skip to content

Latest commit

 

History

History
11 lines (7 loc) · 735 Bytes

README.md

File metadata and controls

11 lines (7 loc) · 735 Bytes

commonsense-part-whole

Dataset for the EMNLP 2019 paper

Do Nuclear Submarines Have Nuclear Captains? A Challenge Dataset for Commonsense Reasoning over Adjectives and Objects

James Mullenbach, Jonathan Gordon, Nanyun Peng and Jonathan May

The dataset consists of part-whole relations along with an adjective describing the whole. The task is to infer the relationship between the adjectives and the parts.

  • data/annotated contains the dataset in lexical format (i.e. just the whole, part, and adjective words), and in NLI format (with retrieved premises and templated hypotheses).
  • mturk/ contains the prompt templates and scripts used to create the AMT tasks.