HASOC Subtrack

The HASOC subtrack focuses on automatically identifying hate speech and offensive content in various languages, aiming to improve online content moderation and mitigate the spread of harmful language. Current research emphasizes the development of robust classification models, often employing transformer architectures and techniques like span detection to pinpoint the specific hateful elements within text. This work is significant for its potential to enhance online safety and provide valuable benchmark datasets and evaluation methods for the broader natural language processing and computer vision communities.