A Comparative Study of Deep Learning Models for Hate Speech Detection on Social Media
A Comparative Study of Deep Learning Models for Hate Speech Detection on Social Media
Keywords:
Analisis Sentimen Multimodal, Deteksi Ujaran Kebencian, Pembelajaran Mendalam, VisualBERT, Media Sosial Indonesia, Integrasi Teks dan Gambar, Pemrosesan Bahasa Alami, Visi KomputerAbstract
This study presents a comparative analysis of three deep learning models BERT, CNN-LSTM, and VisualBERT for multimodal hate speech detection on Indonesian social media platforms, specifically Twitter and Instagram. Using a qualitative approach, the research evaluates the models' performance in classifying hate speech expressed through a combination of textual and visual data. The dataset comprises 5,000 multimodal entries reflecting diverse hate speech themes such as religion, ethnicity, gender, and political identity. Evaluation metrics include accuracy, precision, recall, and F1-score. Results show that VisualBERT outperforms the other models with an accuracy of 90.2%, precision of 88.7%, recall of 87.9%, and F1-score of 88.3%, highlighting the effectiveness of simultaneous text and image integration. However, challenges remain in detecting subtle forms of hate speech like sarcasm and irony that require deeper cultural and contextual understanding. The study underscores the importance of multimodal approaches and culturally adapted datasets for effective hate speech detection in Indonesian social media. Findings contribute to advancing automated content moderation technologies and inform policy development aimed at fostering safer online environments.
References
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of Deep
Bidirectional Transformers for Language Understanding. NAACL-HLT, 4171–4186.
Fortuna, P., & Nunes, S. (2018). A Survey on Automatic Detection of Hate Speech in Text.
ACM Computing Surveys, 51(4), 1–30.
Gao, L., & Huang, R. (2020). Visual-Linguistic BERT: A Simple and Performant Baseline for
Vision and Language. arXiv preprint arXiv:2008.12377.
Kim, Y. (2014). Convolutional Neural Networks for Sentence Classification. Proceedings of
the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 1746–1751.
Kurniawan, H., & Wahyudi, S. (2019). Analysis of Hate Speech on Indonesian Social Media.
International Journal of Social Science and Humanity, 9(7), 205–211.
Mozafari, M., Farahbakhsh, R., & Crespi, N. (2020). A BERT-based Transfer Learning
Approach for Hate Speech Detection in Online Social Media. Complex Networks and Their Applications VIII, 928–940.
Mathew, B., Dutt, R., Goyal, P., & Mukherjee, A. (2021). Spread of Hate Speech in Online
Social Media. ACM Transactions on Social Computing, 4(2), 1–35.
Pratama, M., & Sari, P. (2021). Hate Speech Detection in Indonesian Twitter Using Machine
Learning and Deep Learning Techniques. Journal of Informatics and Computing, 5(2), 45–55.
Rahman, M., & Shafique, M. (2020). Multimodal Hate Speech Detection in Social Media.
IEEE Access, 8, 135578–135590.
Ptaszynski, M., & Blunsom, P. (2021). Multimodal Sentiment Analysis: A Survey of
Datasets, Tasks, and Methods. arXiv preprint arXiv:2107.08094.
Sun, C., Myers, A., Vondrick, C., Murphy, K., & Schmid, C. (2019). VideoBERT: A Joint
Model for Video and Language Representation Learning. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 7464–7473.
Waseem, Z., & Hovy, D. (2016). Hateful Symbols or Hateful People? Predictive Features for
Hate Speech Detection on Twitter. Proceedings of the NAACL Student Research Workshop, 88–93.
Xu, J., Choi, J., & Cardie, C. (2020). VisualBERT: A Simple and Performant Baseline for
Vision and Language. arXiv preprint arXiv:1908.03557.
Yang, Z., Qi, P., Zhang, S., Bengio, Y., Cohen, W., Salakhutdinov, R., & Manning, C. D.
(2019). XLNet: Generalized Autoregressive Pretraining for Language Understanding. Advances in Neural Information Processing Systems, 32.
Zhang, Z., Robinson, D., & Tepper, J. (2018). Detecting Hate Speech on Twitter Using a
Convolution-GRU Based Deep Neural Network. European Semantic Web Conference, 745–760.
