Improving Unsupervised Commonsense Reasoning Using Knowledge-Enabled Natural Language Inference
Improving Unsupervised Commonsense Reasoning Using Knowledge-Enabled Natural Language Inference
复制标题
使用知识支持的自然语言推理改进无监督常识推理
DOI：
10.18653/v1/2021.findings-emnlp.420
复制
发表时间：
2021
影响因子：
9.3
通讯作者：
Yongmei Liu
中科院分区：
文献类型：
--
作者：
Canming Huang;Weinan He;Yongmei Liu
最近基于预训练语言模型的方法在常识推理方面表现出了很强的有监督性能。然而，它们依赖昂贵的数据注释和耗时的培训。因此，我们将重点放在无监督常识推理上。我们展示了使用一个通用框架自然语言推理(NLI)来解决各种常识推理任务的有效性。通过利用从大型NLI数据集的迁移学习，并注入来自原子2020和概念网等常识源的关键知识，我们的方法在两个常识推理任务WinoWHY和CommonsenseQA上实现了最先进的无监督性能。进一步的分析证明了多类别知识的好处，但关于数量和反义词的问题仍然具有挑战性。
Recent methods based on pre-trained language models have shown strong supervised performance on commonsense reasoning. However, they rely on expensive data annotation and time-consuming training. Thus, we focus on unsupervised commonsense reasoning. We show the effectiveness of using a common framework, Natural Language Inference (NLI), to solve diverse commonsense reasoning tasks. By leveraging transfer learning from large NLI datasets, and injecting crucial knowledge from commonsense sources such as ATOMIC 2020 and ConceptNet, our method achieved state-of-the-art unsupervised performance on two commonsense reasoning tasks: WinoWhy and CommonsenseQA. Further analysis demonstrated the benefits of multiple categories of knowledge, but problems about quantities and antonyms are still challenging.
DOI：
10.18653/v1/2020.emnlp-main.411
发表时间：
2020-10
期刊：
--
影响因子：
--
作者：
Jianguo Zhang;Kazuma Hashimoto;Wenhao Liu;Chien-Sheng Wu;Yao Wan;Philip S. Yu;R. Socher;Caiming Xiong
通讯作者：
Jianguo Zhang;Kazuma Hashimoto;Wenhao Liu;Chien-Sheng Wu;Yao Wan;Philip S. Yu;R. Socher;Caiming Xiong