IOParlspeech: A Dataset of Over 600,000 Statements on International Organizations in National Parliamentary Debates

IOParlspeech is a dataset of over 600,000 statements on international organizations (IOs) in national parliamentary debates between 1990 and 2018. The data covers six countries: Austria, Canada, Germany, United Kingdom, United States, and New Zealand. It is generated from the Parlspeech V2 dataset (Rauh and Schwalbach 2021) and from parsed speeches in the US Congress and Canadian House of Commons (Gentzkow et al. 2018, Beelen et al. 2017). Keyword in context is applied to find mentions of IOs (full names and acronyms), as well as extensive validation to remove false positives. Meta-data include information on date, speaker, party, IO mentioned, amongst others. The article 'International Organizations in National Parliamentary Debates', published at the Review of International Organizations provides illustrations of the use of IOParlspeech and introduces the data. Please see also the data documentation on Dataverse.IOParlspeech是一款收录1990至2018年各国议会辩论中涉及国际组织（International Organizations，IO）的超过60万条发言的数据集。该数据集覆盖奥地利、加拿大、德国、英国、美国与新西兰六个国家。其数据源自Parlspeech V2数据集（Rauh与Schwalbach，2021），以及美国国会与加拿大下议院的经解析辩论发言文本（Gentzkow等，2018；Beelen等，2017）。研究团队采用上下文关键词法识别国际组织的全称与缩写形式，并通过多轮严格校验剔除误报条目。数据集元数据涵盖发言日期、发言人、所属党派、提及的国际组织等多项信息。发表于《国际组织评论》（Review of International Organizations）的论文《各国议会辩论中的国际组织》，不仅展示了IOParlspeech数据集的应用示例，还对该数据集进行了系统介绍。更多细节可参阅Dataverse平台上的数据集文档。