Abstract:

Phishing is to capture sensitive information such as username and password of online users through the impersonation of a trustworthy website by internet fraudsters. This information theft, which first appeared in 1996 and is described the victims as a fish, is called phishing. This problem, known as website phishing, is becoming more and more critical as the number of online payments increases, in the electronic banking and electronic commerce sectors. The purpose of this work is to separate legitimate websites from phishing websites using a database that UCI has published. There are 9 feature vectors and 1353 samples in the used database. The feature vector consists of information about the destination URL address. Samples are labeled in three classes: Legitimate, Suspicious, and Phishing. Collected features are converted into category values. For example: Legitimate, Suspicious, and Phishing are listed as 1, 0, and -1, respectively. The purpose of the study is to determine whether the visited website is phishing by giving the feature vector as input to known machine learning methods. As a result of the experimental studies made, all samples were classified with 90% accuracy.

Keywords: