IIE-NLP-NUT at SemEval-2020 Task 4: Guiding PLM with Prompt Template Reconstruction Strategy for ComVE

TL;DR

This paper presents a prompt-based input reconstruction strategy for commonsense validation and explanation tasks, formalizing them as multiple-choice questions and achieving top-tier accuracy in SemEval-2020.

Contribution

The authors introduce a novel prompt template and input reconstruction method to improve PLM performance on commonsense validation and explanation tasks.

Findings

Achieved third place in both subtasks with 96.4% and 94.3% accuracy.

Significant performance improvement over baseline systems.

Effective formalization of subtasks as multiple-choice question answering.

Abstract

This paper introduces our systems for the first two subtasks of SemEval Task4: Commonsense Validation and Explanation. To clarify the intention for judgment and inject contrastive information for selection, we propose the input reconstruction strategy with prompt templates. Specifically, we formalize the subtasks into the multiple-choice question answering format and construct the input with the prompt templates, then, the final prediction of question answering is considered as the result of subtasks. Experimental results show that our approaches achieve significant performance compared with the baseline systems. Our approaches secure the third rank on both official test sets of the first two subtasks with an accuracy of 96.4 and an accuracy of 94.3 respectively.