Scholarlink

<p>Racism, encompassing various forms of discrimination based on race, religion, ethnicity, language, and other identity factors, represents a growing global problem, particularly in a digital age where technology dominates communication. With the rise of social networks, racist content continues to spread more quickly and become harder to control, making automated detection of racist content essential. This work presents a comprehensive and up-to-date multilingual survey on racism detection, focusing on three major languages: French, English, and Arabic. The survey covers studies from 2017 to 2025, highlighting significant progress in the field, including the application of deep learning methods, the emergence of transformer-based models, and advancements in multilingual transfer learning. In addition, we review common techniques used for hate speech detection and show how they can be adapted for racism detection, as racism is often viewed as a specific type of hate speech. This study aims to give an overview of existing work, current challenges, and future research directions in racism detection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Detecting racism in the digital age: a survey of datasets and algorithms

  • 期刊论文

  • 发表日期：

    2026年3月9日

查看全文
  • Ikram El Miqdadi,
  • Soufiane Hourri,
  • Assia Hayati,
  • Yassine Namir,
  • Fatima Zahra El Idrysy,
  • Jamal Kharroubi

摘要

Racism, encompassing various forms of discrimination based on race, religion, ethnicity, language, and other identity factors, represents a growing global problem, particularly in a digital age where technology dominates communication. With the rise of social networks, racist content continues to spread more quickly and become harder to control, making automated detection of racist content essential. This work presents a comprehensive and up-to-date multilingual survey on racism detection, focusing on three major languages: French, English, and Arabic. The survey covers studies from 2017 to 2025, highlighting significant progress in the field, including the application of deep learning methods, the emergence of transformer-based models, and advancements in multilingual transfer learning. In addition, we review common techniques used for hate speech detection and show how they can be adapted for racism detection, as racism is often viewed as a specific type of hate speech. This study aims to give an overview of existing work, current challenges, and future research directions in racism detection.