A Semantic-based Method for Unsupervised Commonsense Question Answering
This addresses the challenge of improving accuracy and robustness in unsupervised commonsense QA for NLP applications, though it is incremental as it builds on existing generative models.
The paper tackles the problem of unsupervised commonsense question answering by proposing a semantic-based method that generates plausible answers and selects the correct choice based on semantic similarity, achieving state-of-the-art results on four benchmark datasets and showing stronger robustness against adversarial attacks with less performance drop.
Unsupervised commonsense question answering is appealing since it does not rely on any labeled task data. Among existing work, a popular solution is to use pre-trained language models to score candidate choices directly conditioned on the question or context. However, such scores from language models can be easily affected by irrelevant factors, such as word frequencies, sentence structures, etc. These distracting factors may not only mislead the model to choose a wrong answer but also make it oversensitive to lexical perturbations in candidate answers. In this paper, we present a novel SEmantic-based Question Answering method (SEQA) for unsupervised commonsense question answering. Instead of directly scoring each answer choice, our method first generates a set of plausible answers with generative models (e.g., GPT-2), and then uses these plausible answers to select the correct choice by considering the semantic similarity between each plausible answer and each choice. We devise a simple, yet sound formalism for this idea and verify its effectiveness and robustness with extensive experiments. We evaluate the proposed method on four benchmark datasets, and our method achieves the best results in unsupervised settings. Moreover, when attacked by TextFooler with synonym replacement, SEQA demonstrates much less performance drops than baselines, thereby indicating stronger robustness.