Abstract
As social media platforms offer a medium for opinion expression, social phenomena such as hatred, offensive language, racism, and all forms of verbal violence have increased spectacularly. These behaviors do not affect specific countries, groups, or communities only, extending beyond these areas into people’s everyday lives. This study investigates offensive and hate speech on Arab social media to build an accurate offensive and hate speech detection system. More precisely, we develop a classification system for determining offensive and hate speech using a multi-task learning (MTL) model built on top of a pre-trained Arabic language model. We train the MTL model on the same task using cross-corpora representing a variation in the offensive and hate context to learn global and dataset-specific contextual representations. The developed MTL model showed a significant performance and outperformed existing models in the literature on three out of four datasets for Arabic offensive and hate speech detection tasks.
Highlights
In recent years, the use of the social networks has substantially increased in the Arab world
multi-task learning (MTL)-A-L and MTL-A-T: are MTL models with AraBERT used in the shared part, and open-source Arabic corpora and corpora processing tools (OSACT)-offensive language detection (OFF), OSACT-hate speech (HS), and T-HSAB are used in the specific task part;
MTL-M-L and MTL-M-T: are MTL models with MarBERT covering Maghreb region dialect and modern standard Arabic (MSA) used in the shared part, and OSACT-OFF, OSACT-HS, and T-HSAB are used in the specific task part;
Summary
The use of the social networks has substantially increased in the Arab world. It has allowed more freedom for opinion expression, especially in the political domain. Due to the freedom of speech given to social media users, it has become relatively easy to propagate abusive or hate speech towards individuals, groups, or societies. Online hate speech is characterized as the use of an offensive language, aimed at a specific group of people who share some common trait [1], while social networks have been recognized as a very favorable medium often used for planning and executing hate attack related activities [2]. It is important to detect such cases of cyber-aggression and cyber-bullying in good time [4]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
