Investigating the Performance of Deep Learning in Detecting Phishing Website


  • Noormadinah Allias Universiti Kuala Lumpur
  • Noor Dayana


This project investigates the performance of deep learning in detecting the phishing websites. Phishing website is one of the security threats in the Internet that can disturb any web services. The phishing websites aim to steal user’s information, such as usernames, passwords, credit cards including financial details. It is very difficult to identify the difference between phishing and legitimate website. Therefore, the purpose of this project is to investigate the performance of machine learning in classifying the phishing website by using a deep learning model. The deep learning algorithm to learn the URL features identified in the dataset during the training process. The other purpose is to analysis the performance of the technique and model in term of accuracy, precision and recall. The result was compared between the other two classification models which is Random Forest and Decision Tree. The result of accuracy achieved for deep learning model is higher than the other models. The result from this project shows that the model and technique proposed outperforms in terms of accuracy and low error rate.