jejjr/OpenCaselist

OpenCaselist数据集是一个用于大学和高中辩论比赛的证据集合，是DebateSum数据集的扩展版本，增加了更多的元数据和范围。数据集涵盖了2013-2024年间的Policy、Lincoln Douglas和Public Forum辩论风格的证据。证据经过去重处理，以检测不同辩论中使用的相同证据。数据集还包含了丰富的元数据，如辩论回合信息、证据来源、辩论队伍和学校等。数据集的主要用途包括论证挖掘、论证分类、论证关系检测、论证生成等任务，也可用于抽象和提取式摘要、自然语言推理等任务。数据集由高中和大学辩论选手创建，主要来源于新闻文章、期刊文章、书籍等。数据集中的姓名和学校信息经过匿名化处理，仅包含姓名的前两个字符。

OpenCaselist is a collection of evidence used in collegiate and high school debate competitions. It is a follow-up to the DebateSum dataset, expanding its scope and the amount of metadata collected. The dataset includes evidence from Policy, Lincoln Douglas, and Public Forum debate styles from the years 2013-2024. Evidence was deduplicated to detect the same piece of evidence used in different debates. The dataset also includes rich metadata about debate rounds, evidence sources, debate teams, and schools. The primary uses of the dataset include argument mining, argument classification, argument relation detection, argument generation, and other related tasks. It is also useful for abstractive and extractive summarization, natural language inference, and as a pretraining dataset for many tasks. The dataset was created by high school and collegiate debaters, sourced from news articles, journal articles, books, and other materials. Personal information such as names and schools has been anonymized, with only the first two characters of names included.