HRDsAttack

HRDsAttack数据集由数据挖掘公司创建，专注于检测针对人权捍卫者的攻击。该数据集包含500篇在线新闻文章的众包标注，每篇文章标注了13种不同的事件属性，如攻击类型、地点和受害者信息。数据集通过关键词过滤和目标抽样确保类别平衡，支持人权组织进行历史或持续人权侵犯的回顾性分析。HRDsAttack旨在通过NLP技术快速处理大量新闻文章，提取攻击特征，支持人权领域的研究和应用，特别是全球此类事件影响的有效管理。

The HRDsAttack dataset was developed by a data mining company, with a core focus on detecting attacks targeting human rights defenders. This dataset contains crowdsourced annotations for 500 online news articles, where each article is labeled with 13 distinct event attributes such as attack type, location, and victim information. The dataset ensures class balance through keyword filtering and targeted sampling, enabling human rights organizations to conduct retrospective analyses of historical or ongoing human rights violations. HRDsAttack aims to rapidly process large volumes of news articles using natural language processing (NLP) techniques, extract attack-related features, and support research and applications in the human rights domain, particularly for the effective management of the impacts of such incidents globally.