Phishing Website Detection: An Improved Accuracy through Feature Selection and Ensemble Learning
Alyssa Anne Ubinga, Syukrina Kamiliaa b, Azween B. Abdullaha, Noor Zaman Jhanjhic d, Mahadevan Supramaniama b
- aSchool of Computing and IT, Taylors University, Subang Jaya, Malaysia
- bSchool of Computing and IT , Taylors University , Subang Jaya , Malaysia
- cResearch & Innovation Management Centre , SEGI University , Malaysia
- dResearch & Innovation Management Centre, SEGI University, Malaysia
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
This research focuses on evaluating whether a website is legitimate or phishing. Our research contributes to improving the accuracy of phishing website detection. Hence, a feature selection algorithm is employed and integrated with an ensemble learning methodology, which is based on majority voting, and compared with different classification models including Random forest, Logistic Regression, Prediction model etc. Our research demonstrates that current phishing detection technologies have an accuracy rate between 70% and 92.52%. The experimental results prove that the accuracy rate of our proposed model can yield up to 95%, which is higher than the current technologies for phishing website detection. Moreover, the learning models used during the experiment indicate that our proposed model has a promising accuracy rate.
TL;DRThe experimental results prove that the accuracy rate of the proposed model can yield up to 95%, which is higher than the current technologies for phishing website detection. (Semantic Scholar)