PhishLens XAI_Explainable Multilingual Phishing Detection System

Phishing attacks are a serious cybersecurity problem because they can be found in emails, messages, and URLs written in different languages. This research proposed PhishLens-XAI: Explainable Multilingual Phishing Detection System to solve two main problems: detecting phishing in different languages and explaining why the system gives a particular prediction.   The proposed system uses multilingual text processing, URL features, machine-learning models, and Explainable AI (XAI). It classifies the input as phishing or legitimate and uses SHAP and LIME to show the important features that affected the prediction. The system is evaluated using accuracy, precision, recall, and F1-score, with performance compared across the selected languages.   The main benefit of PhishLens-XAI is that it does not only provide a phishing prediction. It also provides an explanation of the decision. This can help users and security analysts better understand common phishing signs and make better decisions.   However, the performance of the system may be different for different languages because the amount and quality of available data may vary. Future research can improve the system by using larger datasets, adding more languages, and testing the system against new phishing attacks. Overall, PhishLens-XAI provides a simple approach for detecting phishing across multiple languages while also explaining the reasons behind its predictions. It can provide a useful foundation for future research in multilingual and explainable phishing detection.  

[1]  M. C. Calzarossa, P. Giudici, and R. Zieni, “Explainable machine learning for phishing feature detection,” Quality and Reliability Engineering International, vol. 40, no. 1, pp. 362–373, 2024, doi: 10.1002/qre.3411.   [2]  M. C. Calzarossa, P. Giudici, and R. Zieni, “Explainable Machine Learning for Bag of Words-Based Phishing Detection,” in Explainable Artificial Intelligence, 2023.   [3]   F. Greco, G. Desolda, and A. Esposito, “Explaining Phishing Attacks: An XAI Approach to Enhance User Awareness and Trust,” in Proc. ITASEC, 2023.   [4]  Y. L. D. H. Yakandawala et al., “Explainable AI for Transparent Phishing Email Detection,” in Proc. ICATC, 2024.   [5]  D. Staples, S. Hakak, and P. Cook, “A Comparison of Machine Learning Algorithms for Multilingual Phishing Detection,” in Proc. PST, 2023.   [6]  S. Apeke, T. Djiraikinan, and M. Barate, “Phishing detection by Transformers: Adversarial Robustness and Cross-Lingual Generalization,” Indian Journal of Science and Technology, vol. 19, no. 12, 2026. [7]   Shah et al., “Multilingual Phishing Analysis: Feature-Based Scoring and Level Classification,” in Proc. ICSFT, 2026.   [8]  H. K. M. Lai et al., “Explainable Machine Learning Framework for Phishing URL Detection,” in Proc. CSPA, 2026.   [9]    Y. Chai et al., “An Explainable Multi-Modal Hierarchical Attention Model for Developing Phishing Threat Intelligence,” IEEE Trans. Dependable Secure Comput., vol. 19, no. 2, pp. 790–803, 2022.   [10]  H. Pandya et al., “Detection of Phishing Websites and Emails Using Explainable Machine Learning Models,” Journal of Information Systems Engineering and Management, 2025.