PolInterviews
收藏资源简介:
PolInterviews数据集是由德国社会科学莱布尼茨研究所、开姆尼茨大学和曼海姆大学联合创建的一个新颖数据集,包含99个德国高级政治家的公开广播采访,涵盖五种主要采访形式,共计28,146个句子。数据来源于YouTube,经过Whisper模型转录和ECAPA-TDNN模型进行说话者识别,最终通过人工验证确保质量。该数据集旨在研究政治沟通中的议程设置、采访者动态以及政治家的自我展示等问题,适用于定量和定性分析,为政治科学和媒体研究提供了宝贵资源。
The PolInterviews dataset is a novel dataset jointly created by the Leibniz Institute for the Social Sciences in Germany, Chemnitz University, and the University of Mannheim. It contains 99 public broadcast interviews with senior German politicians, covering five main interview formats and totaling 28,146 sentences. The data was sourced from YouTube, transcribed using the Whisper model, with speaker identification performed via the ECAPA-TDNN model, and finally verified manually to ensure data quality. This dataset aims to research topics including agenda-setting in political communication, interviewer dynamics, and politicians' self-presentation, and supports both quantitative and qualitative analysis, providing a valuable resource for political science and media research.

- 1PolInterviews -- A Dataset of German Politician Public Broadcast Interviews德国社会科学莱布尼茨研究所, 开姆尼茨大学, 曼海姆大学 · 2025年



