Improving Phishing Detection Using Attention-Augmented CNN-GRU-RNN Networks

View at publisher Summarise with VIRAT
Authors: Animesh Ranjan, Abha Singh Sardar, Vivek Ranjan, Nityanand Sharma
Cited by: 0
References: 21

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Phishing attacks are a widespread problem online. Attackers use deceptive URLs to trick people into revealing sensitive information when they click on these links. Various methods have been developed to fight these attacks, often relying on machine learning. The effectiveness of machine learning algorithms depends on the data they are trained on, impacting their accuracy and performance. Deep learning, a specific type of machine learning using neural networks, generally offers high accuracy and can learn complex patterns over time. To improve phishing detection, we propose a new hybrid deep learning model. This model combines Convolutional Neural Networks (CNNs), Bidirectional Gated Recurrent Units (BiGRUs), and Recurrent Neural Networks (RNNs), and is further improved with an Attention Mechanism. The CNN part of our model identifies key local patterns within URL text. The BiGRU and RNN layers then analyze the sequence of characters in the URL, understanding relationships from both earlier and later parts of the string. By adding an Attention Mechanism, our model can pinpoint the most suspicious parts of a URL. This greatly improves its ability to tell the difference between safe and malicious links. Our tests show that this combined approach leads to more accurate detection. It also results in a more stable learning process and makes it easier to understand how the model makes its decisions. This makes our model particularly well-suited for fighting the constantly evolving tactics used in phishing attacks.

Fields of Study & Subjects

Spam and Phishing Detection Misinformation and Its Impacts Hate Speech and Cyberbullying Detection

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
902–908
Publication date
21 January 2026
Language
EN