E-ISSN 3041-4849
 
2024, Vol: 1, Issue: 2
1 / 2Current Issue ArchiveAims and ScopeAbstracting & IndexingMost Accessed ArticlesMost Downloaded ArticlesMost Cited Articles

Research Article

Online Publishing Date:
25 / 07 / 2024

 


Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer

Enesi Femi Aminu, Ayobami EKUNDAYO, Shedrack David SARKIBAKA, Oluwaseun Adeniyi OJERINDE, Uchenna Cosmas UGWUOKE.


Abstract
Aim: The research aims to create a new hate-speech detection model by utilizing hybridized method that capture complex contextual linkages within textual data. Hate speech remain a threat to peaceful coexistence of humans in societies especially via open social networks in this current age, presenting grave obstacles to online safety and promoting inclusive environments.
Methods: This is achieved by combining the advantages of BERT attention processes with a context analyzer. Careful data augmentation was carried out utilizing back translation, which is made possible by the deep-translator library, enhancing the dataset's diversity and quantity to guarantee a comprehensive and reliable dataset.
Results: The training of the frozen BERT layer out of the two layers of the model produced a total accuracy of 0.99 on the 20th epoch by identifying the multi-labeled classes of hate speech using the Adam optimizer and softmax. Promising performance is shown by the trained model's assessment metrics, which include a macro precision of 0.79875, a macro recall of 0.71587, and a macro F1-score of 0.74825.
Conclusion: By utilizing the hybridized BERT model, damaging information can be understood holistically as it can identify not only explicit hate speech but also subtle sensitivities and underlying meanings.

Key words: Hate Speech, Context Analyzer, BERT-Attention Mechanism, Natural Language Processing (NLP), Detection Model


 
ARTICLE TOOLS
Abstract
PDF Fulltext
How to cite this articleHow to cite this article
Citation Tools
Related Records
 Articles by Enesi Femi Aminu
Articles by Ayobami EKUNDAYO
Articles by Shedrack David SARKIBAKA
Articles by Oluwaseun Adeniyi OJERINDE
Articles by Uchenna Cosmas UGWUOKE
on Google
on Google Scholar


How to Cite this Article
Pubmed Style

Aminu EF, EKUNDAYO A, SARKIBAKA SD, OJERINDE OA, UGWUOKE UC. Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. J Comp Sci Informatics. 2024; 1(1): 17-32. doi:10.5455/JCSI.20240613103822

Web Style

Aminu EF, EKUNDAYO A, SARKIBAKA SD, OJERINDE OA, UGWUOKE UC. Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. https://www.wisdomgale.com/jcsi/?mno=205657 [Access: August 30, 2024]. doi:10.5455/JCSI.20240613103822

AMA (American Medical Association) Style

Aminu EF, EKUNDAYO A, SARKIBAKA SD, OJERINDE OA, UGWUOKE UC. Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. J Comp Sci Informatics. 2024; 1(1): 17-32. doi:10.5455/JCSI.20240613103822


Vancouver/ICMJE Style

Aminu EF, EKUNDAYO A, SARKIBAKA SD, OJERINDE OA, UGWUOKE UC. Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. J Comp Sci Informatics. (2024), [cited August 30, 2024]; 1(1): 17-32. doi:10.5455/JCSI.20240613103822


Harvard Style

Aminu, E. F., EKUNDAYO, . A., SARKIBAKA, . S. D., OJERINDE, . O. A. & UGWUOKE, . U. C. (2024) Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. J Comp Sci Informatics, 1 (1), 17-32. doi:10.5455/JCSI.20240613103822


Turabian Style

Aminu, Enesi Femi, Ayobami EKUNDAYO, Shedrack David SARKIBAKA, Oluwaseun Adeniyi OJERINDE, and Uchenna Cosmas UGWUOKE. 2024. Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. Journal of Computer Sciences and Informatics, 1 (1), 17-32. doi:10.5455/JCSI.20240613103822


Chicago Style

Aminu, Enesi Femi, Ayobami EKUNDAYO, Shedrack David SARKIBAKA, Oluwaseun Adeniyi OJERINDE, and Uchenna Cosmas UGWUOKE. "Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer." Journal of Computer Sciences and Informatics 1 (2024), 17-32. doi:10.5455/JCSI.20240613103822


MLA (The Modern Language Association) Style

Aminu, Enesi Femi, Ayobami EKUNDAYO, Shedrack David SARKIBAKA, Oluwaseun Adeniyi OJERINDE, and Uchenna Cosmas UGWUOKE. "Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer." Journal of Computer Sciences and Informatics 1.1 (2024), 17-32. Print. doi:10.5455/JCSI.20240613103822


APA (American Psychological Association) Style

Aminu, E. F., EKUNDAYO, . A., SARKIBAKA, . S. D., OJERINDE, . O. A. & UGWUOKE, . U. C. (2024) Hate Speech Detector Based on Hybridized BERT-Attention Mechanism and Context Analyzer. Journal of Computer Sciences and Informatics, 1 (1), 17-32. doi:10.5455/JCSI.20240613103822


Author LoginReviewer LoginAbout PublisherOpen Access PolicyEditorial PoliciesEditorial Review PolicyPeer Review PolicyEditorial & Peer Review ProcessPublication Ethics and Publication Malpractice StatementConflict of Interest PolicyPlagiarism PolicyProtection of Research Participants (Statement On Human And Animal Rights)Privacy PolicyCorrections, Retractions & Expressions of ConcernSelf-Archiving PoliciesDigital Archiving & Preservation PoliciesTerms of UseLicense InformationCopyright Information

About Journal of Computer Sciences and Informatics

Journal of Computer Sciences and Informatics is an international, peer-reviewed, open access journal, providing a platform for advances in basic, translat ... Read more.



For best results, please use Internet Explorer or Google Chrome.

Contact Information

Your comments are very important to us

We welcome your ideas, suggestions, comments, or questions. To reach Editorial Board please use one of the methods provided below.
 

Emails:

info@wisdomgale.com
editorinchief@ejmanager.com

Office Address:
WisdomGale Publishing
14 Rue de Grand-Bigard,
1082 Brussels
Belgium
 
We will respond within 48 to 72 hours.
 
Change of address or personal information? Please visit your account at www.ejmanager.com to update your information.