TY - JOUR
AU - 
AB - Unsupervised commonsense question answer- ing 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 con- ditioned on the question or context. However, such scores from language models can be eas- ily 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 can- Figure 1: Two examples of commonsense question an- didate answers. swering, where the baseline (Pro-A) is oversensitive to In this paper, we present a novel SEmantic- lexical perturbations (SR for synonym replacement and based Question Answering method (SEQA) ST for sentence structure transformation). The scores for unsupervised commonsense question an- from Pro-A and our method for each answer choice are swering. Instead of directly scoring each an- shown in the right columns. The underlined score indi- swer choice, our method first generates a set cates the answer choice selected by a method. of plausible answers with generative models (e.g., GPT-2), and then uses these plausible an- Khashabi et al., 2020; Lin et al., 
TI - A Semantic-based Method for Unsupervised Commonsense Question Answering
JF - Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
DO - 10.18653/v1/2021.acl-long.237
DA - 2021-01-01
UR - https://www.deepdyve.com/lp/unpaywall/a-semantic-based-method-for-unsupervised-commonsense-question-NvJsoeCkq0
DP - DeepDyve
ER -