A Question-Answer Driven Approach to Reveal Affirmative Interpretations from Verbal Negations
This addresses a specific challenge in natural language processing for improving machine understanding of nuanced language, but it is incremental as it builds on existing methods like NLI and T5.
The paper tackled the problem of extracting affirmative interpretations from verbal negations, finding that 67.1% of such negations in a new corpus of 4,472 examples convey an actual event, and state-of-the-art transformers underperformed humans in generating these interpretations.
This paper explores a question-answer driven approach to reveal affirmative interpretations from verbal negations (i.e., when a negation cue grammatically modifies a verb). We create a new corpus consisting of 4,472 verbal negations and discover that 67.1% of them convey that an event actually occurred. Annotators generate and answer 7,277 questions for the 3,001 negations that convey an affirmative interpretation. We first cast the problem of revealing affirmative interpretations from negations as a natural language inference (NLI) classification task. Experimental results show that state-of-the-art transformers trained with existing NLI corpora are insufficient to reveal affirmative interpretations. We also observe, however, that fine-tuning brings small improvements. In addition to NLI classification, we also explore the more realistic task of generating affirmative interpretations directly from negations with the T5 transformer. We conclude that the generation task remains a challenge as T5 substantially underperforms humans.