NECTAR

该数据集是一个问答数据集，每个问题配备了七个可能的答案。在其他信息方面，该数据集被用于带有真实反馈的实验中，实验考虑了随机提问的情况。任务是对答案或段落根据与问题的相关性进行排名。

This dataset is a question answering (QA) dataset, with each question paired with seven candidate answers. It has been employed in experiments involving real-world feedback, where random question sampling is taken into consideration. The core task of this dataset is to rank answers or passages based on their relevance to the corresponding question.