AmbigNQ

AmbigNQ, a dataset covering 14,042 questions from NQ-open, an existing open-domain QA benchmark. We find that over half of the questions in NQ-open are ambiguous. The types of ambiguity are diverse and sometimes subtle, many of which are only apparent after examining evidence provided by a very large text corpus. AMBIGNQ, a dataset with 14,042 annotations on NQ-OPEN questions containing diverse types of ambiguity. We provide two distributions of our new dataset AmbigNQ: a full version with all annotation metadata and a light version with only inputs and outputs.

歧义自然问题数据集（AmbigNQ）是一款基于现有开放域问答（Question Answering，QA）基准NQ-open构建的数据集，共涵盖14042个源自该基准的问题。研究发现，NQ-open中超过半数的问题存在歧义，此类歧义类型多样且往往较为隐晦，其中多数仅在查阅超大规模文本语料库提供的证据后方可显现。AmbigNQ是针对NQ-open中各类歧义问题构建的标注数据集，共包含14042条标注样本。本次发布的AmbigNQ提供两种分发版本：包含全部标注元数据的完整版本，以及仅保留输入与输出内容的轻量版本。