The suggested approach uses shared representation knowledge across five Arabic dialects—Egyptian, Saudi, Levant, Gulf, and Algerian—and is intended to detect and differentiate subtle hate speech patterns using publically accessible datasets from different dialects. To the best of our knowledge, it is the first model to use the unique features of each dialect to identify hate speech while concurrently addressing numerous dialects. Results demonstrate that, in comparison to single-task models, the suggested model significantly advances hate speech identification in the Arabic language. With F1 ratings of 0.98, 0.84, 0.85, 0.76, and 0.80 for the Egyptian, Levant, Saudi, Algerian, and Gulf dialects, respectively, it represented a 14% improvement over earlier studies.https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5101030Share this:FacebookXLike this:Like Loading... Post navigation A survey of textual cyber abuse detection using cutting-edge language models and large language models (arXiv) Ensuring safety in digital spaces: Detecting code-mixed hate speech in social media posts (Data & Knowledge Engineering)