Enhanced Fake News Detection Using AMR-Based Text Augmentation and Adversarial Training
Keywords:
Abstract meaning representation, Adversarial training, BiLSTM, CNN, Fake newsAbstract
Fake news has become a critical challenge due to its rapid dissemination through social media and digital platforms. Conventional deep learning models for fake news detection often face limitations related to insufficient training data and vulnerability to adversarial text manipulation. This study proposes an enhanced fake news detection framework that integrates Abstract Meaning Representation (AMR)-based semantic text augmentation, CLONALG, salience-based AMR graph reduction, and adversarial training. The proposed approach generates diverse and semantically meaningful training samples through concept, relation, focus, and node mutations while preserving the core meaning of the original text. Salience-based graph reduction further removes less important nodes to reduce complexity and noise. The augmented, original, and adversarial samples are classified using BiLSTM and CNN models. Experiments on COVID-19 News, FakeNewsNet, and Kaggle Fake News datasets demonstrate improved accuracy, precision, recall, and F1-score, together with reduced adversarial attack success rates. The framework therefore enhances both fake news detection performance and robustness against textual manipulation.
References
C. Shao, Y. Zi, Y. Deng, H. Liu, C. Zhang, and Y. Ni, “Adversarial Robustness in Text Classification Through Semantic Calibration With Large Language Models,” in 2026 International Conference on Communication Networks and Machine Learning (CNML), IEEE, Jan. 2026, pp. 1371–1375.
G. L. Anand Babu, G. Sekhar Reddy, G. Sravan Kumar, A. Prashanth Rao, N. Nagalakshmi, and S. Vijay Kumar, “A Hybrid Framework for Fake News Detection Using Transformer Models With MASHAP,” International Journal of Data Science and Analytics, vol. 21, no. 1, Dec. 2025.
A. Safaa Mahdi and N. Mezaal Shati, “A Survey on Fake News Detection in Social Media Using Graph Neural Networks,” Journal of Al-Qadisiyah for Computer Science and Mathematics, vol. 16, no. 2, Jun. 2024.
H. Ahmed, I. Traore, S. Saad, and M. Mamun, “Effect of Text Augmentation and Adversarial Training on Fake News Detection,” IEEE Transactions on Computational Social Systems, vol. 11, no. 4, pp. 4775–4789, Aug. 2024.
H. Tang, S. Kamei, and Y. Morimoto, “Data Augmentation Methods for Enhancing Robustness in Text Classification Tasks,” Algorithms, vol. 16, no. 1, p. 59, Jan. 2023.
S. Gong, R. O. Sinnott, J. Qi, and C. Paris, “Fake news detection through graph-based neural networks: A survey,” arXiv preprint arXiv:2307.12639, Jul. 2023.
B. Lakshmidevi, J. T., M. B., and S. E., “Multilingual Fake News Detection Using Multimodal Transformers and Semi-Supervised Machine Learning,” International Conference on Emerging Trends in Artificial Intelligence and Blockchain Technology, 2026, pp. 265-270.
M. Zhao, Y. Zhang, and G. Rao, “Fake News Detection Based on Dual-Channel Graph Convolutional Attention Network,” The Journal of Supercomputing, vol. 80, no. 9, pp. 13250–13271, Mar. 2024.
P. Kantartopoulos, N. Pitropakis, A. Mylonas, and N. Kylilis, “Exploring Adversarial Attacks and Defences for Fake Twitter Account Detection,” Technologies, vol. 8, no. 4, p. 64, Nov. 2020.
J. Wei and K. Zou, “EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks,” Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019, pp. 6382-6388.
A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, and I. Sutskever, “Language models are unsupervised multitask learners,” OpenAI blog, vol. 1, no. 8, p. 9, 2019.
M. Peters et al., “Deep Contextualized Word Representations,” Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), vol. 1, 2018.
A. Radford and T. Salimans, “Improving language understanding by generative pretraining,” OpenAI, 2018.
L. Yu, W. Zhang, J. Wang, and Y. Yu, “SeqGAN: Sequence Generative Adversarial Nets With Policy Gradient,” Proceedings of the AAAI Conference on Artificial Intelligence, Feb. 2017, vol. 31, no. 1.
