A Framework for Vietnamese Email Phishing Detection

Abstract

Currently, the attacks on network information systems are increasing rapidly in number and level of danger. Phishing email distribution method is widely used by hackers today. This is an attacking technique that exploits human behaviors in the system. This type of attack, though it is not too complicated, becomes very effective for attackers if users are unaware of information security and unable to identify phishing emails. This attack is particularly more commonly effective in developing countries where the information security is still overlooked. As a result, email phishing detection problem has become a hot topic for information security researchers. There have been some published methods to detect phishing emails on given email attacking datasets. However, one of the important issues in email phishing detection relates to the language used in emails. Each particular language used in different emails may lead to a different phising detection approach. In this article, a Vietnamese email phishing detection system is investigated. The research includes a feature selection method and a combination of machine learning algorithms to improve the performance of phishing email detection in Vietnamese language. The proposed method is evaluated using two datasets. The first dataset includes phishing emails from Vietnamese collected from Vietnamese volunteers. The second dataset is the widely used English emails as introduced in [16,17]. The experimental results show that our method is applicable for real Vietnamese email phishing detection systems.