Comparative analysis of transformer models and hybrid architectures in phishing content detection tasks
DOI:
https://doi.org/10.15330/itee.2025.2.02Keywords:
phishing content, phishing detection, natural language processing, transformer models, BERT, hybrid neural networks, CNN, LSTM, attention mechanism, text classificationAbstract
The work demonstrates that the use of hybrid neural network architectures based on CNN, LSTM, and the attention mechanism provides higher efficiency in phishing content detection compared to the transformer-based BERT model. The experimental analysis conducted using K-Fold cross-validation and standard evaluation metrics confirmed the advantages of combining local text analysis with long-term context modeling. It was found that the integration of the attention mechanism increases classification accuracy and recall by focusing on the most informative text fragments. The obtained results demonstrate the potential of hybrid approaches for building practical phishing attack detection systems.
References
M. Vijayalakshmi, S. Mercy Shalinie, M. H. Yang, and R. M. U., “Web phishing detection techniques: a survey on the state-of-the-art, taxonomy and future directions,” IET Netw., vol. 9, no. 5, pp. 235–246, Sep. 2020. doi: https://doi.org/10.1049/iet-net.2020.0078.
M. A. Adebowale, K. T. Lwin, and M. A. Hossain, “Deep Learning with Convolutional Neural Network and Long Short-Term Memory for Phishing Detection,” in 2019 13th Int. Conf. Softw., Knowl., Inf. Manage. Appl. (SKIMA), Island of Ulkulhas, Maldives, Aug. 26–28, 2019. IEEE, 2019. doi: https://doi.org/10.1109/skima47702.2019.8982427.
S. Priya, D. Gutema, and S. Singh, “A Comprehensive Survey of Recent Phishing Attacks Detection Techniques,” in 2024 5th Int. Conf. Innovative Trends Inf. Technol. (ICITIIT), Kottayam, India, Mar. 15–16, 2024. IEEE, 2024. doi: https://doi.org/10.1109/icitiit61487.2024.10580446.
S. Atawneh and H. Aljehani, “Phishing Email Detection Model Using Deep Learning,” Electronics, vol. 12, no. 20, p. 4261, Oct. 2023. doi: https://doi.org/10.3390/electronics12204261.
U. Daniel, E. Bartholomew, and F. Egbono, “Phishing URL Attack Detection using Logistic Regression and Convolutional Neural Network,” Int. J. Comput. Appl., vol. 187, no. 1, pp. 8–14, May 2025. doi: https://doi.org/10.5120/ijca2025924611.
B. Vishnupriya, B. Vikas, and M. D. Choudhry, “Hybrid Deep Learning Framework for ECG-Based Arrhythmia Detection Using CNN, Bi-LSTM, and Transformer,” in 2025 IEEE 4th World Conf. Appl. Intell. Comput. (AIC), GB Nagar, Gwalior, India, Jul. 26–27, 2025. IEEE, 2025, pp. 628–632. doi: https://doi.org/10.1109/aic66080.2025.11212074.
A. A. Tawil, L. Almazaydeh, D. Qawasmeh, B. Qawasmeh, M. Alshinwan, and K. Elleithy, “Comparative Analysis of Machine Learning Algorithms for Email Phishing Detection Using TF-IDF, Word2Vec, and BERT,” Comput., Mater. & Continua, pp. 1–10, 2024. doi: https://doi.org/10.32604/cmc.2024.057279.
M. E. Maurer, "Phishload," URL: https://www.medien.ifi.lmu.de/team/max.maurer/files/phishload .
S. Luo, Y. Gu, X. Yao, and W. Fan, “Research on Text Sentiment Analysis Based on Neural Network and Ensemble Learning,” Revue d'Intell. Artificielle, vol. 35, no. 1, pp. 63–70, Feb. 2021. doi: https://doi.org/10.18280/ria.350107.
K. Zhou, Y. Zhou, W. X. Zhao, and J.-R. Wen, “Learning to Perturb for Contrastive Learning of Unsupervised Sentence Representations,” IEEE/ACM Trans. Audio, Speech, Lang. Process., pp. 1–10, 2023. doi: https://doi.org/10.1109/taslp.2023.3304485.
M. Abdar et al., “A review of uncertainty quantification in deep learning: Techniques, applications and challenges,” Inf. Fusion, vol. 76, pp. 243–297, Dec. 2021. doi: https://doi.org/10.1016/j.inffus.2021.05.008.
Downloads
Published
How to Cite
Issue
Section
