ISCA-IUB/HateSpeechAndBias

该数据集是印第安纳大学当代反犹太主义研究所（ISCA）关于针对亚裔、黑人、犹太人、拉丁裔和穆斯林的偏见信息的研究项目成果。数据集包含2020、2021和2022年从Twitter档案中抓取的包含关键词Asians, Blacks, Jews, Latinos, and Muslims的实时消息。每年每个关键词随机抽取600条推文，包括转发。这些推文由2022年和2023年秋季Gunther Jikeli教授课程Researching White Supremacism and Antisemitism on Social Media的本科生进行标注。标注内容包括偏见和揭露偏见的问题，使用1到5的偏见量表进行评分。标注结果通过75%多数投票确定。数据集分为2022年和2023年两个队列，分别标注了2020、2021和2021、2022年的数据。数据集以CSV文件格式提供，包含推文ID、用户名、推文文本、创建日期、偏见标签、揭露偏见标签、关键词和标注队列等信息。

This dataset is the research outcome of a project conducted by the Institute for the Study of Contemporary Antisemitism (ISCA) at Indiana University, focusing on biased information targeting Asians, Blacks, Jews, Latinos, and Muslims. The dataset contains real-time tweets crawled from Twitter archives in 2020, 2021 and 2022 that include the keywords Asians, Blacks, Jews, Latinos and Muslims. For each keyword in each year, 600 tweets (including retweets) are randomly sampled. These tweets were annotated by undergraduate students in Professor Gunther Jikeli's course "Researching White Supremacism and Antisemitism on Social Media" during the fall semesters of 2022 and 2023. The annotation covers bias and bias-exposing content, with ratings assigned using a 1-to-5 bias scale. Annotation results are determined via 75% majority voting. The dataset is divided into two cohorts: the 2022 cohort and the 2023 cohort, which annotated data from 2020, 2021 and 2021, 2022 respectively. The dataset is provided in CSV format, containing fields such as tweet ID, username, tweet text, creation date, bias label, bias-exposing label, keyword, and annotation cohort.

原始信息汇总

数据集概述

数据来源

• 数据集来源于印第安纳大学的一项研究项目，专注于Twitter上针对种族和宗教少数群体的偏见信息。

数据收集

• 收集时间：2020年至2022年。
• 数据内容：通过关键词“Asians, Blacks, Jews, Latinos, and Muslims”从Twitter档案中抓取的所有实时消息。