A Collective Inference Framework for Phishing URL Detection Using Loopy Belief Propagation.

In this paper, we developed PhiURL, a phishing URL detection framework leveraging Loopy Belief Propagation, graph-based relational modelling, and machine learning-based probabilistic inference. The proposed framework integrates lexical, domain-based, TF-IDF, and statistical URL features to improve phishing detection accuracy and classification stability. URLs are represented as interconnected graph nodes, where similarity-based pairwise relationships enable collective inference across the graph structure. Experimental results demonstrate that integrating Random Forest classification with Loopy Belief Propagation significantly improves phishing detection performance compared to traditional machine learning approaches.   The proposed framework achieved high accuracy, precision, recall, and F1-score while maintaining effectiveness on large-scale phishing datasets. Furthermore, the real-time prediction capability and probabilistic graph inference mechanism make the proposed PhiURL framework suitable for scalable cybersecurity and phishing detection applications.

 1. Efficient Phishing URL Detection Using Graph Based Machine Learning and Loopy Belief Propagation.     2. Wenye Guo, Qun Wang, Hao Yue, Haijian Sun, and Rose Qingyang Hu , “Efficient Phishing URL Detection Using Graph-Based Machine Learning and Loopy Belief Propagation,” arXiv preprint, Jan. 2025   3. A systematic literature review on phishing website detection techniques Safi and S. Singh, “A Systematic Literature Review on Phishing Website Detection   4. Techniques,” J. King Saud Univ. Comput. Inf. Sci., vol. 35, no. 2, pp. 590–611, Jan. 2023.   5. A Survey of Machine Learning-Based Solutions for Phishing Website Detection   6. L. Tang and Q. H. Mahmoud., “A Survey of Machine Learning-Based Solutions for Phishing Website Detection,” Information, vol. 3, no. 3, p. 34, 2023. https://www.mdpi.com/2504- 4990/3/3/34   7. Phishing Website Detection Based on Multidimensional Features Driven by Deep Learning   8. P. Yang, G. Zhao and P. Zeng, “Phishing Website Detection Based on Multidimensional Features Driven by Deep Learning,” IEEE Access, 2018. https://ieeexplore.ieee.org/abstract/document/ 8610190   9. A graph-theoretic approach for the detection of phishing webpages   10. C. C. L. Tan, K. L. Chiew, K. S. C. Yong, S. N. Sze, J. Abdullah, and Y. Sebastian, “A Graph-Theoretic Approach for the Detection of Phishing Webpages,” Computers & Security, vol. 92, 2020.