Detection and Prediction of Phishing Websites using Classification Mining Techniques

Abstract

Phishing is serious web security problem that involves mimicking legitimate websites to deceive online users in order to steal their sensitive information. Phishing can be seen as a typical classification problem in data mining where the classifier is constructed from large number of website's features. There are high demands on identifying the best set of features that when mined the predictive accuracy of the classifiers is enhanced. This paper investigates features selection aiming to determine the effective set of features in terms of classification performance. We compare two known features selection method in order to determine the least set of features of phishing detection using data mining. Experimental tests on large number of features data set have been done using Information Gain and Correlation Features set methods. Further, two data mining algorithms namely C4.5 and IREP have been trained on different sets of selected features to show the pros and cons of the feature selection process. We have been able to identify new knowledge in the forms of rules that show vital correlations among significant features.

Understanding phishers' strategies of mimicking uniform resource locators to leverage phishing attacks: A machine learning approach

“…Al‐diabat, Mofleh 13 proposed a feature selection approach aiming to determine the effective set of features in‐terms of improving the performance of the classification. In this research, they have used two feature selection methods such as IG and symmetrical uncertainty (SU) to detect a small set of correlation among features.…”

“…In all previous research 11,13,14,17,24 work has been focused on improving the phishing attack classification rate, identify the best classifier to identify phishing attack, address the different types of phishing detection approaches and feature selection of the phishing URL to reduce the dimension of the phishing dataset. But they failed to identify strategies or techniques, cyber‐criminals used to mimic URLs to manipulate humans.…”

“…But it takes the visitor to a phishing website without the user's knowledge. However, they have used a set of common features like lengthy URL, anchor URL, suffix and prefix special characters, irregular URL, etc 13 . to mimic these kind of phishing URLs' techniques.…”

“…In this, several features are extracted from the phishing URL to classify either fraudulent or legitimate. In the past literature Waleed Ali 16 proposed a phishing detection approach based on supervised learning with wrapper features selection, Mofleh‐Al‐diabat 13 used classification mining techniques for detecting and predicting phishing website, Neda Abdelhamid 17 proposed an approach for phishing detection based on the associative classification (AC) and Rami M. Mohammed 18 used self‐structuring neural network for predicting phishing website. However, previous research has failed to identify the techniques or strategies used by cyber‐criminals (ie, phishers) to mimic legitimate URLs 19 .…”

An Ensemble Model for Identification of Phishing Website

“…K. Rajitha et al (2016) [5] have analyzed the malicious detection problems .They have offered a survey of the malicious website detection techniques using various phishing method. M. Al-diabat et al (2016) [6] have investigated features selection aiming to determine the effective set of features in terms of classification performance. They compare two known features selection method in order to determine the least set of features of phishing detection using data mining.…”