Algorithm Evaluation for Classification “Phishing Website” Using Several Classification Algorithms

Abstract

Phishing websites are a fooling technique by making victims as if they are accessing legitimate sites. Data mining is a technique for extracting hidden information in order to benefit more from existing data. Data mining is the process of discovering regularity, patterns, and relationships in large datasets. In this study, data mining will be used to determine the effect of feature selection on algorithm C4.5 and CART on phishing website dataset. From the tests that have been done the effect of feature selection on the phishing website, dataset proved to overcome the longer computational time. From the performance measurement of both algorithms that have been done, CART algorithm has a higher accuracy value than the algorithm C4.5 with an accuracy of 94.4%, while the algorithm C4.5 has an accuracy of 94.3%, so it can be concluded that CART algorithm has better performance value compared with the C4.5 algorithm.