A HYBRID ANN-PSO APPROACH FOR EMAIL SPAM DETECTION
Salahahaldeen Duraibi1 and Nawaf A Almolhis2
1Department of Computer and Network Engineering, Jazan University, Jazan 82822-6649, Saudi Arabia
2Department of Computer science, Jazan University, Jazan 82822-6649, Saudi Arabia
Sduraibi@jazanu.edu.sa1 and firstname.lastname@example.org
In today’s world, technology extends into all areas of human life and emails is one of these technologies that expand communication platforms with suitable and inexpensive manner. Market organizations and advertising use these low-cost platforms to spread their desired information in the form of spam. To protect Internet users from spam danger, various strategies, tools, and techniques are presented. Therefore, this paper introduces a hybrid model, namely ANN-PSO which combines Artificial Neural Network (ANN) and Particle Swarm Optimization (PSO) for spam detection. The PSO is employed to pick most important features to be then used as inputs to the ANN model. The developed ANN-PSO model is assessed using several evaluation measurements and compared to ANN, k-Nearest Neighbor (KNN), Logistic Regression (LR) and Support Vector Machine (SVM) models. The results show that proposed ANN-PSO approach got a promising outcomes and achieved better performance than the other comparative models.