A systematic literature review of hate speech identification on Arabic Twitter data: research challenges and future directions

• DOI: 10.7717/peerj-cs.1966
• Publication Date: 2024-04-02T07:15:26Z
• Authors: Ali Alhazmi; Rohana Mahmud; Norisma Idris; Mohamed Elhag Moh...; Christopher Eke
• Keywords: Identification

The automatic speech identification in Arabic tweets has generated substantial attention among academics the fields of text mining and natural language processing (NLP). quantity studies done on this subject experienced significant growth. This study aims to provide an overview field by conducting a systematic review literature that focuses hate identification, particularly language. goal is examine research trends offer guidance researchers highlighting most published between 2018 2023. addresses five specific questions concerning types used, categories, classification techniques, feature engineering performance metrics, validation methods, existing challenges faced researchers, potential future directions. Through comprehensive search across nine academic databases, 24 met predefined inclusion criteria quality assessment were identified. findings revealed existence many linguistic varieties used Twitter, with modern standard (MSA) being prominent. In machine learning categories are technique for identification. result also shows different techniques indicates N-gram CBOW techniques. F1-score, precision, recall, accuracy identified as metric. method train/test split method. Therefore, can serve valuable enhancing efficacy their models investigations. Besides, algorithm development, policy rule regulation, community management, legal ethical consideration other real-world applications be reaped from research.