ExaASC

ExaASC是一个针对阿拉伯语的目标基础立场检测语料库，由伊朗的Exa公司创建。该数据集包含9566条数据，主要从Twitter上收集，涉及多种类型的目标，如人物、事件和实体。数据集的创建过程包括从约80,000条源推文中筛选出争议性和辩论性较强的推文，并通过专业标注者进行目标和立场的标注。ExaASC的应用领域主要集中在社交媒体分析，特别是政治讨论中的立场检测，旨在通过分析推文及其回复来确定对特定目标的立场。

ExaASC is a target-based stance detection corpus dedicated to the Arabic language, developed by Exa Corporation of Iran. This dataset comprises 9,566 samples, primarily collected from Twitter and covering a wide range of target categories including people, events, and entities. The construction of ExaASC involved filtering out highly controversial and debate-driven tweets from approximately 80,000 source tweets, followed by annotation of targets and their corresponding stances by professional annotators. The primary application scenarios of ExaASC focus on social media analytics, particularly stance detection in political discussions, with the goal of determining the stance towards specific targets by analyzing both tweets and their replies.