ReviewQA

ReviewQA是一个基于酒店评论的问答数据集，由NAVER LABS Europe创建。该数据集包含超过500,000个自然语言问题，覆盖100,000条酒店评论，旨在评估模型对评论内容的理解和推理能力。数据集通过众包和回译技术生成，包含8种不同的推理任务类型，用于测试模型在不同方面的理解和回答能力。ReviewQA的应用领域主要集中在机器阅读理解，特别是解决自然语言处理中的理解和推理问题。

ReviewQA is a question answering dataset built upon hotel reviews, developed by NAVER LABS Europe. This dataset comprises over 500,000 natural language questions spanning 100,000 hotel reviews, with the core objective of evaluating models' understanding and reasoning capabilities regarding review content. The dataset is generated through crowdsourcing and back-translation techniques, and it encompasses 8 distinct reasoning task types to assess models' comprehension and response capabilities across diverse dimensions. The primary application scope of ReviewQA lies in machine reading comprehension, particularly in addressing comprehension and reasoning challenges in natural language processing.