Phishing URL detection by using artificial neural network with PSO

Abstract

Phishing is a serious threat in the virtual world and it has affected the information system. Phishing is mostly conducted via email, other methods such as instant messaging and social networking sites as Facebook twitter. In this paper, we proposed a method to classify the Uniform Resource Locator (URL) into Phishing URL or Non phishing URL. Artificial neural network has been trained by using particle swarm optimization (PSO) to classify URLs to improve the performance of ANN. The proposed model has been run on different ratio of learning and different activation function on number of hidden layer, output layer. Accuracy and RMSE criteria have been selected for evaluating the artificial neural network with particle swarm optimization (ANN_PSO) model. ANN_PSO model performs better training in terms of accuracy with respect to Back Propagation Neural Network (BPNN).