Transforming Question Answering Datasets Into Natural Language Inference Datasets

  • 2018-09-11 00:38:25
  • Dorottya Demszky, Kelvin Guu, Percy Liang
  • 0

Abstract

Existing datasets for natural language inference (NLI) have propelledresearch on language understanding. We propose a new method for automaticallyderiving NLI datasets from the growing abundance of large-scale questionanswering datasets. Our approach hinges on learning a sentence transformationmodel which converts question-answer pairs into their declarative forms.Despite being primarily trained on a single QA dataset, we show that it can besuccessfully applied to a variety of other QA resources. Using this system, weautomatically derive a new freely available dataset of over 500k NLI examples(QA-NLI), and show that it exhibits a wide range of inference phenomena rarelyseen in previous NLI datasets.

 

Quick Read (beta)

loading the full paper ...