Collecting Diverse Natural Language Inference Problems for Sentence Representation Evaluation
Collecting Diverse Natural Language Inference Problems for Sentence Representation Evaluation
复制标题
DOI：
10.18653/v1/d18-1007
复制
发表时间：
2018-04
期刊：
影响因子：
--
通讯作者：
Adam Poliak;Aparajita Haldar;Rachel Rudinger;J. E. Hu;Ellie Pavlick;Aaron Steven White;Benjamin Van Durme
中科院分区：
文献类型：
--
作者：
Adam Poliak;Aparajita Haldar;Rachel Rudinger;J. E. Hu;Ellie Pavlick;Aaron Steven White;Benjamin Van Durme
我们提供了大量不同的自然语言推理（NLI）数据集，有助于深入了解句子表示捕获不同类型推理的能力。该集合是将来自 7 种语义现象的 13 个现有数据集重新转换为通用 NLI 结构的结果，总共产生了超过 50 万个标记的上下文假设对。我们将我们的集合称为 DNC：多样化自然语言推理集合。 DNC 可在 https://www.decomp.net 上在线获取，并且随着时间的推移，随着更多资源的重新分配和从新来源添加，该 DNC 将不断增长。
We present a large-scale collection of diverse natural language inference (NLI) datasets that help provide insight into how well a sentence representation captures distinct types of reasoning. The collection results from recasting 13 existing datasets from 7 semantic phenomena into a common NLI structure, resulting in over half a million labeled context-hypothesis pairs in total. We refer to our collection as the DNC: Diverse Natural Language Inference Collection. The DNC is available online at https://www.decomp.net, and will grow over time as additional resources are recast and added from novel sources.