Multilingual and Multi-Aspect Hate Speech Analysis (SSRN)

The majority of current hate speech analysis research focuses on single-task categorization and monolingual datasets. The scholars present a new multilingual, multi-aspect hate speech analysis dataset in this work and use it to evaluate the latest multilingual multitask learning methods. The authors test our dataset in several categorization contexts and investigate how our annotations might improve the identification and classification of hate speech.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5148442