•  
  •  
 

Document Type

Article

Keywords

Phishing attack detection, Deep Learning, Binary Particle Swarm Optimization, BPSO, Feature Selection

Abstract

Phishing attacks manipulate users to disclose critical information, resulting in cybersecurity risks. Traditional phishing detection algorithms usually have large false positive rates and poor feature selection, degrading performance. This paper presents an optimized phishing detection framework that integrates binary particle swarm optimization (BPSO)-based feature selection (FS) with deep learning models. Six deep learning architectures were evaluated on the selected feature subset to identify the most effective model for accurate phishing classification. BPSO was used to select suitable attributes on a public Kaggle dataset with 10,000 samples, comprising phishing and legitimate website data with 48 attributes. NumDots, UrlLength, IpAddress, and NoHttps were selected among the 25 features chosen. BPSO was chosen because it effectively reduces feature dimensionality while preserving crucial attributes that enhance classification accuracy. The BPSO optimally selects relevant phishing-related attributes, improving model efficiency and reducing computational complexity. The BPSO technique optimally selects the most relevant features, reducing dataset dimensionality by 48% while maintaining high classification performance. We used six DL models—MLP, 1D-CNN, RNN, LSTM, GRU, and DNN—to test the specified characteristics. The experimental results demonstrate that the DNN model outperforms the other methods through 99.63% accuracy, 99.74% precision, 99.54% recall, and an AUC of 0.9999.

Download

Share

COinS