Related Experiment Video
Updated: May 19, 2025
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
1.3K
A hybrid super learner ensemble for phishing detection on mobile devices.
Routhu Srinivasa Rao1,2, Cheemaladinne Kondaiah3, Alwyn Roshan Pais3
1Department of Computer Science and Engineering, Gandhi Institute of Technology and Management, Visakhapatnam, Andhra Pradesh, 530045, India.
Scientific Reports
|May 15, 2025
Summary
Phish-Jam is a new mobile app that effectively detects phishing websites using a hybrid super learner ensemble. This advanced machine learning model achieves high accuracy, protecting users from cyber threats on mobile devices.
Area of Science:
- Cybersecurity
- Machine Learning
- Mobile Security
Background:
- Phishing attacks pose a significant threat in the digital age, targeting sensitive user information through deceptive websites.
- Existing anti-phishing techniques have limitations, including vulnerability to zero-day attacks and unsuitability for mobile devices with constrained resources.
Purpose of the Study:
- To develop a novel, efficient, and robust phishing detection model for mobile devices.
- To address the limitations of current anti-phishing methods, particularly on resource-constrained mobile platforms.
Main Methods:
- Proposed Phish-Jam, a hybrid super learner ensemble model for mobile phishing detection.
- Extracted URL features using handcrafted features, transformer-based text embeddings, and Deep Learning architectures.
- Combined predictions from diverse Machine Learning algorithms for website classification.
Main Results:
- The Phish-Jam model achieved high performance metrics: 98.93% accuracy, 99.15% precision, 97.81% MCC, and 99.07% F1 Score.
- Demonstrated advantages including fast computation, language independence, and robustness against malware downloads.
Conclusions:
- The proposed Phish-Jam model offers a significant advancement in mobile phishing detection.
- The super learner ensemble approach provides an effective solution for identifying and mitigating phishing threats on mobile devices.
