Twitter Hate Speech Detection: A Systematic Review of Methods, Taxonomy Analysis, Challenges, and Opportunities

Abstract

Hate speech detection has substantially increased interest among researchers in the domain of natural language processing (NLP) and text mining. The number of studies on this topic has been growing dramatically. Thus, the purpose of this analysis is to develop a resource that consists of an outline of the approaches, methods, and techniques employed to address the issue of Twitter hate speech. This study can be used to aid researchers in the development of a more effective model for future studies. This review focused on studies published over the past eight years, i.e., from 2015 to 2022. This systematic search was carried out in December 2020 and updated in July 2022. Ninety-one articles published within the mentioned period met the set criteria and were selected for this review. From the evaluation of these works, it is clear that a perfect solution has yet to be found. To conclude, this paper focused on presenting an in-depth understanding of current perspectives and highlighted research opportunities to boost the quality of hate speech detection systems. In turn, this helps social networking services that seek to detect hate messages generated by users before they are posted, thus reducing the risk of targeted harassment.

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

“…On the other hand, deep learning techniques eliminate the need for handcrafted features. Deep learning has gained significant popularity for HS identification in Arabic Twitter data since 2017 ( Badjatiya et al, 2017 ), primarily due to its capacity to research classification appropriate to data representations ( Husain, 2020 ; Mansur, Omar & Tiun, 2023 ). Well-known deep learning techniques include CNNs and LSTM networks ( Duwairi, Hayajneh & Quwaider, 2021 ).…”

Multimodal Hate Speech Detection in Memes Using Contrastive Language-Image Pre-Training

“…The rise of online hateful messages in recent times has emerged as a pressing societal concern, posing significant harm to both individuals and the community at large [2]. These "Hateful Memes" represent a particularly vexing challenge.…”

Hate speech detection in social media: Techniques, recent trends, and future challenges

“…Numerous interpretations of hate speech are available, but a prevailing thread among them is the utilization of language that either targets or encourages violence against specific groups characterized by distinct attributes. Even with the progress in HSD techniques, assessments frequently prioritize the detection of non-hate content, rather than the specific task of identifying and categorizing hateful content (Mansur et al, 2023). In HSD, ML techniques-which include ensemble approaches, conventional ML, and deep learning (DL) have made significant strides.…”

“…The systematic review covers computational methods for HSD, discussing existing challenges and highlighting areas for further research opportunities. The aim is to enhance HSD methods, aligning with the goals of governments and social media companies to create a more user-friendly environment by curbing hate speech (Mansur et al, 2023). This article presents significant contributions concerning the answers to various research questions presented in Table 5 of Section 3.…”

“…The systematic review covers computational methods for HSD, discussing existing challenges and highlighting areas for further research opportunities. The aim is to enhance HSD methods, aligning with the goals of governments and social media companies to create a more user‐friendly environment by curbing hate speech (Mansur et al, 2023). This article presents significant contributions concerning the answers to various research questions presented in Table 5 of Section 3. We have undertaken an investigation encompassing research papers published over 5 years, ranging from 2018 to 2023, with a specific emphasis on HSD within the realm of social media, spanning across diverse platforms. We have demonstrated a multitude of global incidents stemming from hate speech in the years 2022–2023, spanning various regions around the world. We have conducted an exhaustive examination and synthesis of the existing studies within this particular research domain. We categorized HSD methods into five distinct categories: ML, DL, Ensemble, Graph Neural Networks (GNN), and Graph Convolutional Networks (GCN). We have provided an overview of various datasets that are accessible for HSD. We presented prominent challenges in research and anticipated potential future trends. The remaining content of this document is structured into seven sections.…”