arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.09099cs.CLcs.AIcs.LG
收藏

面向下游任务的句子级自然语言推理

Looking Beyond Sentence-Level Natural Language Inference for Downstream Tasks

College of Information and Computer Sciences, University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校信息与计算机科学学院);IBM Research(IBM研究院)

Anshuman Mishra, Dhruvesh Patel, Aparna Vijayakumar, Xiang Li, Pavan Kapanipathi, Kartik Talamadupula

摘要:

近年来,自然语言推理(NLI)任务获得了极大的关注,新的数据集和模型在其上实现了接近人类水平的性能。然而,NLI的全部承诺-特别是它学习的知识应该推广到其他下游NLP任务-尚未实现。在本文中,我们从两个下游任务的镜头:问答(QA)和文本摘要研究这个未实现的承诺。我们推测,NLI数据集和这些下游任务之间的关键区别在于前提的长度;并且从现有的QA数据集创建新的长前提NLI数据集是训练真正可推广的NLI模型的有希望的途径。我们验证了我们的猜想,通过显示竞争力的结果,QA的任务,并获得最好的报告结果的任务,检查事实的正确性摘要。

英文摘要:

In recent years, the Natural Language Inference (NLI) task has garnered significant attention, with new datasets and models achieving near human-level performance on it. However, the full promise of NLI -- particularly that it learns knowledge that should be generalizable to other downstream NLP tasks -- has not been realized. In this paper, we study this unfulfilled promise from the lens of two downstream tasks: question answering (QA), and text summarization. We conjecture that a key difference between the NLI datasets and these downstream tasks concerns the length of the premise; and that creating new long premise NLI datasets out of existing QA datasets is a promising avenue for training a truly generalizable NLI model. We validate our conjecture by showing competitive results on the task of QA and obtaining the best reported results on the task of Checking Factual Correctness of Summaries.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑