Phishing Attack Detection using Feature Selection Techniques

7 Pages Posted: 12 Jul 2019 Last revised: 30 Sep 2019

Date Written: May 18, 2019


The cyber security problems are increasing nowadays due to the growth of internet world wide. As phishing links have different features it becomes very difficult to mitigate different security issues. In phishing attack attacker creates a replica of existing link or webpage to fool the user to get access to the personal information such as credit card or debit card passwords and try to convince the victim that the message originates from the proper source. As phishing links involves around 20 to 40 features it becomes somehow difficult to address each feature so features selection algorithms selects important features which are relevant to the attack. The proposed system has selected important features of URL using machine learning classifiers and feature selection algorithms. The system has tested accuracy against different combinations of classifiers and feature selection algorithms from which Random Forest algorithm as a binary classifier and ReliefF algorithm which is feature selection algorithm have performed better than other combinations. To detect phishing attacks various techniques which are machine learning based, Anti-phishing tools are discussed in related work of the paper.

Keywords: Feature Selection, ReliefF Algorithm, Random Forest Algorithm

JEL Classification: Y60

Suggested Citation

Joshi, Anirudha and Pattanshetti, Prof. Tanuja R, Phishing Attack Detection using Feature Selection Techniques (May 18, 2019). Proceedings of International Conference on Communication and Information Processing (ICCIP) 2019, Available at SSRN: or

Anirudha Joshi (Contact Author)

CoE, Shivajinagar, Pune ( email )


Prof. Tanuja R Pattanshetti

CoE, Shivajinagar, Pune ( email )


Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics