Toxic Comment
Toxic online comments, encompassing hate speech, harassment, and other forms of abusive language, are a significant concern, with research focusing on automated detection and mitigation. Current efforts utilize various deep learning architectures, including transformer models like BERT and its variants, along with convolutional and recurrent neural networks, to classify toxicity and identify its targets across multiple languages and modalities (text, images, video). This research is crucial for improving online safety, informing content moderation strategies, and understanding the societal impact of online toxicity, particularly its disproportionate effects on vulnerable groups.
Papers
November 11, 2024
September 25, 2024
September 19, 2024
September 8, 2024
August 21, 2024
July 17, 2024
July 12, 2024
July 8, 2024
May 31, 2024
May 29, 2024
April 2, 2024
March 25, 2024
January 23, 2024
December 13, 2023
November 8, 2023
November 3, 2023
November 1, 2023