Authors: Harichandana B S S, Sumit Kumar, Manjunath Bhimappa Ujjinakoppa, Barath Raj Kandur Raja

Published on: February 06, 2024

Impact Score: 8.52

Arxiv code: Arxiv:2402.04173

Summary

• What is new: A novel, lightweight on-device pipeline named COPS for real-time detection of smishing and phishing attacks on mobile devices.
• Why this is important: The rapid increase in cybercrime targeting mobile devices, specifically through smishing attacks, and the inefficacy of databases of malicious URLs due to their short lifespan.
• What the research proposes: COPS uses a Disentangled Variational Autoencoder to identify features of fraudulent messages and URLs, providing real-time alerts without relying on databases of known malicious URLs.
• Results: Achieved an accuracy of 98.15% for smishing detection and 99.5% for URL phishing detection, with very low false negative and false positive rates, outperforming previous models.

Technical Details

Technological frameworks used: Disentangled Variational Autoencoder

Models used: nan

Data used: Open datasets

Potential Impact

Cybersecurity solution providers, mobile phone manufacturers, e-commerce platforms, and potentially any company involved in digital communication or transactions with customers.

Want to implement this idea in a business?

We have generated a startup concept here: GuardianLink.