Predicting phishing websites based on self-structuring neural network

Abstract

-Internet has become an essential component of our everyday social and financial activities. Nevertheless, internet-users may be vulnerable to different types of web-threats which may cause financial damages, identity theft, loss of private information, brand reputation damage and loss of customer's confidence in ecommerce and online banking. Phishing is considered as a form of web-threats that is defined as the art of impersonating a website of an honest enterprise aiming to obtain confidential information such as usernames, passwords and social security number. So far, there is no single solution that can capture every phishing attack. In this article, we proposed an intelligent model for predicting phishing attacks based on Artificial Neural Network "ANN" particularly self-structuring neural networks. Phishing is a continuous problem where features significant in determining the type of webpages are constantly changing. Thus, we need to constantly improve the network structure in order to cope with these changes. Our model solves this problem by automating the process of structuring the network and shows high acceptance for noisy data, fault tolerance and high prediction accuracy. Several experiments were conducted in our research, the number of epochs differs in each experiment. From the results, we find that all produced structures have high generalization ability.

Application of Feature Engineering for Phishing Detection

“…In this research, we list 16 frequently used features in literatures [8], [21], [23]- [25] and one new feature, and make use of them to perform feature engineering.…”

“…-IP address ( f 1 ): almost all the literatures [8], [21], [23]- [25] [24], [25] followed and used this rule to define the value of this feature. However, this rule was highly affected by the samples.…”

Phishing Website Detection based on Supervised Machine Learning with Wrapper Features Selection

“…The success of phishing website detection techniques mainly depends on recognizing phishing websites accurately and within an acceptable timescale [2], [4]. Many conventional techniques based on fixed black and white listing databases have been suggested to detect phishing websites.…”

“…Mohammad et al [9] suggested rule-based data mining classification techniques with 17 different features to distinguish phishing from legitimate websites. Mohammad et al [4] proposed an intelligent model for predicting phishing attacks based on self-structuring neural networks. Abdelhamid et al [1] developed an approach called Multi-Label Classifier based Associative Classification (MCAC) to detect phishing websites.…”

A classification rules mining method based on dynamic rules' frequency

“…Common applications of classification are medical diagnoses (RameshKumar et al, 2013), phishing detection , fraud detection (Whitten and Frank, 2005), etc. Some of the main classification approaches developed in data mining include Decision Trees (DT) (Quinlan, 1993), Neural Network (NN) (Mohammad et al, 2013), Associative Classification (AC) (Thabtah, et al, 2004) (Thabtah, 2005), and Rule Induction (RI) (Cendrowska, 1987). The latter two approaches extract classifiers that contain human interpretable rules in the form "If-Then", which explain their wide spread applicability.…”