Central AsianResearch Index

USING MACHINE LEARNING TO IDENTIFY HATE SPEECH ON SOCIAL MEDIA

This research work employs a comprehensive examination of various machine learning and deep learning methods to tackle the significant issue of detecting cyberbullying within social media platforms. This study meticulously evaluates these methods through established industry metrics such as F-measure, AUC-ROC, precision, accuracy, and recall, aiming to determine the effectiveness of these approaches. The results demonstrate the effectiveness of deep learning models, particularly the bidirectional long-short-term memory (BiLSTM) architecture, which consistently outperforms alternative methods across various categorization tasks. Confusion matrices and graphical representations offer deeper understanding of the model's performance, highlighting the remarkable ability of the BiLSTM-based model to accurately detect and classify instances of cyberbullying. The findings highlight the importance of advanced neural network architectures in recognizing the complexity of hateful and objectionable content online. This study provides valuable insights for fostering early detection and prevention of cyberbullying, thereby enhancing the safety and inclusivity of online communities. Future studies may explore real-time detection systems, hybrid techniques, or the integration of complementary elements to advance and improve innovative technology in addressing this critical social issue.

Сведения

Опубликовано
2025 · Т. 137, № 2, С. 229-239
Язык
Английский
ISSN
1609-1817 · 2790-5802
Издатель
JSC Academy of Logistics and Transport