CheGeKa

该数据集基于“是什么？在哪里？何时？”这款游戏中的问题，包含了需要逻辑推理和世界知识才能回答的具有挑战性的开放式问题。此外，所报告的评价指标为F1分数。该数据集的规模包括29,376个训练实例和416个测试实例，任务类型为开放式问题回答。

This dataset comprises challenging open-ended questions drawn from the trivia game *What? Where? When?*, which require logical reasoning and general world knowledge for correct answering. Additionally, the reported evaluation metric for this dataset is the F1 score. The dataset includes 29,376 training instances and 416 test instances, targeting the open-ended question answering task.