Leveraging Affirmative Interpretations from Negation Improves Natural Language Understanding
Leveraging Affirmative Interpretations from Negation Improves Natural Language Understanding
复制标题
DOI：
10.48550/arxiv.2210.14486
复制
发表时间：
2022-10
期刊：
影响因子：
--
通讯作者：
Md Mosharaf Hossain;Eduardo Blanco
中科院分区：
文献类型：
--
作者：
Md Mosharaf Hossain;Eduardo Blanco
否定在许多自然语言理解任务中构成了挑战。受理解否定语句通常需要人类推断肯定解释这一事实的启发，在本文中，我们表明这样做有利于三个自然语言理解任务的模型。我们提出了一个自动化的过程来收集对句子的否定和肯定的解释，导致超过150,000对。实验结果表明，利用这些对有助于（a）T5生成肯定的解释，从否定在以前的基准，和（B）基于罗伯塔的分类解决自然语言推理的任务。我们还利用我们的配对来构建一个即插即用的神经生成器，该神经生成器给出否定的陈述，生成肯定的解释。然后，我们将预训练的生成器合并到基于RoberTa的分类器中进行情感分析，并表明这样做可以改善结果。最重要的是，我们的建议不需要任何手动操作。
Negation poses a challenge in many natural language understanding tasks. Inspired by the fact that understanding a negated statement often requires humans to infer affirmative interpretations, in this paper we show that doing so benefits models for three natural language understanding tasks. We present an automated procedure to collect pairs of sentences with negation and their affirmative interpretations, resulting in over 150,000 pairs. Experimental results show that leveraging these pairs helps (a) T5 generate affirmative interpretations from negations in a previous benchmark, and (b) a RoBERTa-based classifier solve the task of natural language inference. We also leverage our pairs to build a plug-and-play neural generator that given a negated statement generates an affirmative interpretation. Then, we incorporate the pretrained generator into a RoBERTa-based classifier for sentiment analysis and show that doing so improves the results. Crucially, our proposal does not require any manual effort.