chiyuanhsiao/text_replay_trivia_qa-audio

该数据集包含了问题及其相关实体页面信息和搜索结果，旨在用于问答系统的研究。每个问题都附带有一个唯一的ID和来源信息。实体页面信息包括文档来源、文件名、标题和维基百科上下文。搜索结果则包括描述、文件名、排名、标题、URL和搜索上下文。答案部分包含了多个字段，如别名、标准化别名、匹配的维基实体名称、标准化匹配的维基实体名称、标准化值、类型和值。此外，还有一个字段是用于存储预测答案文本的。

The dataset consists of questions and their related entity page information and search results, designed for research on question answering systems. Each question is accompanied by a unique ID and source information. The entity page information includes document source, filename, title, and Wikipedia context. The search results consist of description, filename, rank, title, URL, and search context. The answer section contains multiple fields such as aliases, normalized aliases, matched Wikipedia entity name, normalized matched Wikipedia entity name, normalized value, type, and value. Additionally, there is a field for storing predicted answer text.