Arabic Fake News Detection Using Deep Learning and Transformer-Based Models
DOI:
https://doi.org/10.31642/JoKMC/2018/130210Keywords:
Arabic Fake News Detection, AraNews, AraBERT, Deep Learning, Natural Language Processing, CNN, BiLSTM, Transformer ModelsAbstract
With the emergence of digital news platforms and social media, Arabic News content is spreading throughout erroneously. Hence, automatic fake news detection is an important task in natural language processing due to the fact that Arabic suffers from rich morphology, orthographic variations, a great amount of dialects diversity together with code-switching. This study provides a comparative experimental analysis of Arabic fake news detection models and finds the best model architecture on a common dataset and evaluation protocol.
Utilizing balanced Arabic real and fake news articles, the AraNews dataset was used to conduct the experiments. The task was looked at as a classic classification problem with two labelled classes, Real and Fake. The tested models are SVM with TF-IDF features, LSTM, GRU, BiLSTM, CNN-LSTM the AraBERT + 1D-CNN and AraBERT + 2D-CNN. Arabic text was normalized, cleaned, tokenized and vectorized (using TF-IDF, word embeddings or AraBERT contextual embeddings depending on the model type).
The outcome We can see that the best results are in AraBERT + 2D-CNN model, for the accuracy is up to 97.8% and got F1-score →0.98 That said, this discussion does not take the position that deep learning is always best. We discuss our findings in terms of the size of the datasets, how complex or simple a model is, located part on how humans represent context information with an immediate comparison to recent Arabic fake news detection experiments. The results demonstrate that transformer-based contextual embeddings offer a substantial boost for Arabic news classification, whereas conventional and recurrent models are still solid baselines to consider in low-resource or resource-constrained environments.
Downloads
References
[1] A. A. N. Al-Rabeeah and F. Saeed, "Data privacy model for social media platforms," 2017 6th ICT International Student Project Conference (ICT-ISPC), Johor, Malaysia, 2017, pp. 1-5, doi: 10.1109/ICT-ISPC.2017.8075361.
[2] A. Albalawi, R. Ghazzawi, and Y. Alghamdi, “Multimodal Arabic rumor detection model using text and images,” Applied Sciences, vol. 11, no. 22, 2021.
[3] A. Alkhair, H. Alrumaih, and S. Alhumoud, “Arabic fake news detection in YouTube comments,” Proc. 2nd Int. Conf. on Computer Applications & Information Security (ICCAIS), 2019.
[4] F. Alawadh, A. Alotaibi, and H. Alshehri, “Arabic fake news detection using decision tree and random forest,” Int. J. of Advanced Computer Science and Applications (IJACSA), vol. 11, no. 12, 2020.
[5] A. Amoudi, H. Hassan, and M. Ezzeldin, “Arabic fake news detection during COVID-19 using LSTM and BiLSTM,” Proc. Int. Conf. on Computational Linguistics (COLING), 2021.
[6] A. Antoun, F. Baly, and H. Hajj, “AraBERT: Transformer-based model for Arabic NLP,” Proc. Workshop on NLP for Semitic Languages, 2020.
[7] J. Abonizio, M. Garcia, and P. Rosso, “Multilingual fake news detection with feature extraction and machine learning,” Information Processing & Management, vol. 58, no. 6, 2021.
[8] S. Aphiwongsophon and P. Chongstitvatana, “A comparison of machine learning techniques for fake news detection on Twitter,” Proc. Int. Conf. on Advances in Information Technology (IAIT), 2018.
[9] R. Bharadwaj and C. Shao, “Semantic RNN for fake news detection,” Proc. Int. Conf. on Computational Linguistics (COLING), 2018.
[10] H. Himdi, S. Alzahrani, and A. Albathan, “Fake news detection in Arabic during Hajj season using machine learning approaches,” J. of Information Security and Applications, vol. 55, 2020.
[11] M. Hussein, A. Antoun, and H. Baly, “AraBERT for Arabic fake news detection on Twitter,” Proc. 4th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT), 2020.
[12] Hashim, M. M., Al-Hilali, A. A., Safi, M. G. A., Salah, O. R., & Nahi, A. A., “Sentence reduction based on automatic document text summarization and crossbred framework,” 2023 First International Conference on Advances in Electrical, Electronics and Computational Intelligence (ICAEECI), pp. 1-10, IEEE, Oct. 2023.
[13] P. Nakov, A. Ritter, S. Rosenthal, V. Stoyanov, and F. Sebastiani, “Fake news detection: Current challenges and future research directions,” ACL Workshop on Natural Language Processing, 2019.
[14] M. Sorour and M. Abdelkader, “Arabic fake news detection using CNN-LSTM architecture,” Proc. Int. Conf. on Arabic Computational Linguistics (ACLing), 2020.
[15] Y. Wang, J. Ma, and H. Gao, “EANN: Event adversarial neural networks for multimodal fake news detection,” Proc. 24th ACM SIGKDD Int. Conf. on Knowledge Discovery & Data Mining, 2018.
[16] Y. Wang et al., “SemSeq4FD: Semantic sequence model for fake news detection,” Information Sciences, vol. 512, pp. 345–360, 2020.
[17] M. Wotaifi and M. Dhannoon, “Fake news detection in Arabic using CNN-LSTM with AraNews dataset,” Journal of Theoretical and Applied Information Technology (JATIT), vol. 99, no. 7, 2021.
[18] H. Himdi, S. Alzahrani, and A. Albathan, “Fake news detection in Arabic language using machine learning techniques,” Arabian Journal for Science and Engineering, 2022, doi: 10.1007/s13369-021-06449-y.
[19] A. Wotaifi and B. Dhannoon, “Arabic fake news detection using deep learning models on the AraNews dataset,” Baghdad Science Journal, 2023.
[20] M. M. Fouad, S. F. Sabbeh, and W. Medhat, “Arabic fake news detection using deep learning,” Computers, Materials & Continua, vol. 71, no. 2, pp. 2695–2713, 2022.
[21] S. Alyoubi et al., “The detection of fake news in Arabic tweets using deep learning,” Applied Sciences, vol. 13, no. 14, Art. no. 8209, 2023, doi: 10.3390/app13148209.
[22] W. Antoun, F. Baly, and H. Hajj, “AraBERT: Transformer-based model for Arabic language understanding,” in Proc. 4th Workshop on Open-Source Arabic Corpora and Processing Tools, Marseille, France, 2020, pp. 9–15.
[23] M. E. Almandouh et al., “Ensemble-based high-performance deep learning models for fake news detection,” Scientific Reports, vol. 14, Art. no. 76286, 2024, doi: 10.1038/s41598-024-76286-0.
[24] W. Antoun, F. Baly, and H. Hajj, “AraBERTv2: Improved pretrained language model for Arabic,” arXiv preprint arXiv:2101.01785, 2021.
[25] Wotaifi, A., & Dhannoon, B. (2023).
Arabic fake news detection using deep learning models on AraNews dataset.
Baghdad Science Journal.
https://bsj.uobaghdad.edu.iq/index.php/BSJ/article/view/7427
[26] Fouad, M. M., Sabbeh, S. F., & Medhat, W. (2022).
Arabic fake news detection using deep learning models.
Computers, Materials & Continua (CMC).
https://www.techscience.com/cmc/v71n2/45790
[27] Alyoubi, K. H., et al. (2023).
Arabic fake news detection using MARBERT and deep learning.
Applied Sciences (MDPI), 13(14), 8209.
https://doi.org/10.3390/app13148209
[28] W. Antoun, F. Baly, and H. Hajj, “AraBERT: Transformer-based model for Arabic language understanding,” arXiv preprint arXiv:2003.00104, 2020.
[29] Almandouh, M., et al. (2024).
Ensemble-based high-performance deep learning models for fake news detection.
Scientific Reports (Nature), 14, Article 76286.
https://doi.org/10.1038/s41598-024-76286-0.
[30] Antoun, W., Baly, F., & Hajj, H. (2021).
AraBERTv2: Improved pretrained language model for Arabic.
arXiv preprint arXiv:2101.01785.
Downloads
Published
Issue
Section
Categories
License
Copyright (c) 2026 Abdullah Abdulabbas Nahi Alrabeeah, Mohammed Mahdi Hashim , Amna Kadhim Ali
This work is licensed under a Creative Commons Attribution 4.0 International License.
which allows users to copy, create extracts, abstracts, and new works from the Article, alter and revise the Article, and make commercial use of the Article (including reuse and/or resale of the Article by commercial entities), provided the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, indicates if changes were made and the licensor is not represented as endorsing the use made of the work.
How to Cite
