ProvocationProbe
收藏资源简介:
ProvocationProbe是一个旨在探索引发仇恨言论与一般仇恨言论区别的数据集,由印度理工学院鲁尔分校、莱布尼茨社会科学研究所和塔尔图大学共同创建。该数据集包含约27,000条从Twitter收集的推文,涵盖了九个全球争议事件,涉及种族、政治和宗教等多个主题。数据集的创建过程包括从Twitter收集数据、使用Google Perspective API进行预处理和手动标注。该数据集主要用于研究引发仇恨言论的特征及其与非引发仇恨言论的区别,旨在帮助识别和缓解在线仇恨言论的传播。
ProvocationProbe is a dataset dedicated to exploring the distinction between provocative hate speech and general hate speech, jointly created by Indian Institute of Technology Roorkee, Leibniz Institute for Social Sciences, and University of Tartu. This dataset contains approximately 27,000 tweets collected from Twitter, covering nine globally controversial events involving multiple themes such as race, politics, and religion. The dataset's construction process includes data collection from Twitter, preprocessing using Google Perspective API, and manual annotation. It is primarily utilized to study the characteristics of provocative hate speech and its differences from non-provocative hate speech, with the goal of aiding the identification and mitigation of the spread of online hate speech.

- 1ProvocationProbe: Instigating Hate Speech Dataset from Twitter印度理工学院鲁尔分校, 莱布尼茨社会科学研究所, 塔尔图大学 · 2024年



