Semi-supervised Conditional GAN for Simultaneous Generation and Detection of Phishing URLs: A Game theoretic Perspective

Abstract

Spear Phishing is a type of cyber-attack where the attacker sends hyperlinks through email on well-researched targets. The objective is to obtain sensitive information such as name, credentials, credit card numbers, or other crucial data by imitating oneself as a trustworthy website. According to a recent report, phishing incidents nearly doubled in frequency in 2020. In recent times, machine learning techniques have become the standard for defending against such attacks. Many augmentations have been made for improving the existing architectures, such as Convolutional Networks, Recurrent Networks, and Generative Adversarial Networks. However, these architectures were designed with only defense in mind. Moreover, the attacker's perspective and motivation are absent while creating such training and deployment pipelines. To address this, we need a gametheoretic approach to understand the rational decision-making process of the attacker (Hacker) and the defender (Phishing URL detector). We propose a Conditional Generative Adversarial Network for real-time phishing URL detection. Additionally, we train our architecture in a semi-supervised manner to distinguish between adversarial and real examples, along with detecting malicious and benign URLs. We also design two games between the attacker and defender in training and deployment settings by utilizing the game-theoretic perspective. Our experiments confirm that the proposed architecture surpasses recent state-of-theart architectures for phishing URLs detection.

Artificial-Intelligence-Enabled Intelligent 6G Networks

“…Once utilities converge to steady points, they indicate equilibrium strategies. Moreover, we utilize GAN and GDM to generate another two sets of new images, simulating two new games 13 . Following these implementation details, the performance of equilibrium strategies can be evaluated under new generated game scenarios, with the strategy (150, 75, 75) serving as the baseline for the average scheme.…”

“…For instance, deep learning-based Spear Phishing attack defense schemes often neglect attacker models, potentially undermining defense efficacy. To address such a concern,[13] utilizes GAN to design games between the attacker and defender in training and deployment phases. Numerical results have shown that the proposed scheme can realize 0.9552 Area Under Curve (AUC) on Phishing uniform resource locator (URL) classification performance, while baselines Support Vector Machine (SVM), CNN, and LSTM are with 0.8638, 0.9251, and 0.9471 AUC values, respectively.…”

“…https://github.com/Yaoqi-Yang97/GAI4Game13 Note we first utilize GAN and GDM to generate 200 synthetic images as initial dataset at one time, representing action sets of the original game. Then, we create two new games by changing their action sets, i.e., generating another two synthetic datasets for new game scenarios-generated scenarios 1 and 2.…”

A neuro-game model for analyzing strategies in the dynamic interaction of participants of phishing attacks

“…As a result, customers who thought they had bought bitcoin (BTC) and tried to access their funds found that their money had simply disappeared [7]. Spear phishing on CCE is a type of cyber attack in which attackers send hyperlinks, for example, using email, to well-researched targets [8], [9]. The purpose of such attacks is to obtain sensitive information by imitating trusted websites.…”