Phishing Website Detection Paper Published at ICCIT 2025
Our paper Advancing Cybersecurity: A Study on Phishing Website Detection Using Machine Learning has been published at the 2025 28th International Conference on Computer and Information Technology (ICCIT), IEEE.
About the Paper
Phishing attacks remain one of the most prevalent cybersecurity threats, often leading to financial loss and identity theft. Traditional detection mechanisms such as blacklists or heuristic-based filters are insufficient against zero-day phishing URLs and adversarial manipulation. This paper proposes a robust ensemble-based phishing website detection system focusing exclusively on URL-based lexical and structural features.
Key Contributions
- Ensemble framework combining XGBoost, LightGBM, CatBoost, and TabPFN with soft voting
- Focuses exclusively on URL-based lexical and structural features — no page content required
- SHAP-based interpretability for transparent and explainable predictions
- SelectKBest feature selection, StandardScaler normalization, and SMOTE for class balancing
- Suitable for real-time phishing detection deployment
Results
| Metric | Score |
|---|---|
| Accuracy | 93.3% |
| Recall | 0.99 |
Citation
S. N. Sakib, O. Jyoti, and R. K. Kundu, “Advancing Cybersecurity: A Study on Phishing Website Detection Using Machine Learning,” in Proc. 2025 28th International Conference on Computer and Information Technology (ICCIT), IEEE, 2025, pp. 3530–3535.