MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202621051009 A) filed by Balwant Kumar; Abhishek Singh; Asish Singh; Aniket Koratwad; Prof. Dipali Khairnar; Dr. Moresh Madhukar Mukhedkar; Dr. Vivek Patil; Prof. Vishal Patil; and Dr. Shweta Koparde on April 21, 2026, for A Hybrid Multi-Layered Framework For Real-Time Phishing Website Detection Using Bio-Inspired Optimization And Flask.
Inventors include Balwant Kumar; Abhishek Singh; Asish Singh; Aniket Koratwad; Prof. Dipali Khairnar; Dr. Moresh Madhukar Mukhedkar; Dr. Vivek Patil; Prof. Vishal Patil; and Dr. Shweta Koparde.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: Phishing attacks continue to pose a significant threat to online users by exploiting fraudulent websites to steal sensitive information. This study proposes a Hybrid Multi-Layered Framework for Real-Time Phishing Website Detection that integrates Binary Particle Swarm Optimization (BPSO) and ensemble learning. BPSO was employed for intelligent feature selection to reduce dimensionality and enhance classification efficiency. A soft-voting ensemble combining machine learning and deep learning classifiers, such as XGBoost and Convolutional Neural Networks, was implemented and deployed via a Flask-based REST API for real-time inference. The experimental results demonstrate that the optimized ensemble achieves a peak accuracy of 99.12% and a recall of 98.75%. Crucially, BPSO reduced the end-to-end processing latency to 88.4 ms, representing a 75% improvement over the unoptimized frameworks and providing a scalable solution for real-time protection.
Disclaimer: Curated by HT Syndication.
