There has been an increasing number of fake e-Business websites created and used, which have resulted in rising financial loss for online consumers and businesses. Therefore, developing effective approaches to detecting phishing websites is essential to mitigating the possibility of being victimized by those sites and minimizing financial loss and risks. In this research, we propose a novel classification model for automatically detecting Chinese phishing e-Business websites. By extending previous research and incorporating unique characteristics of Chinese e-Business websites, our model consists of feature vectors of both the URL and content of a Website. We have trained and evaluated the proposed model with roughly 900 Chinese e-Business websites using four different classification algorithms. Results show that among those four algorithms, the Sequential Minimal Optimization (SMO) algorithm performs the best. To examine the impact of individual features in the model on detection accuracy, we further conducted a sensitivity analysis to identify the most influential features, which helps make the classification model more parsimonious. The findings of this research provide several research and practical insights into the development of anti-phishing solutions.
Jiang, Hansi; Zhang, Dongsong; and Yan, Zhijun, "A Classification Model for Detection of Chinese Phishing E-Business Websites" (2013). PACIS 2013 Proceedings. 152.