A Hybrid Model to Detect Phishing-Sites Using Supervised Learning Algorithms

Since last decades, online technologies have revolutionized the modern computing world. However, as a result, security threats are increasing rapidly. A huge community is using the online services even from chatting to banking is done via online transactions. Customers of web technologies face various security threats and phishing is one of the most important threat that needs to be address. Therefore, the security mechanism must be enhanced. The attacker uses phishing attack to get victims credential information like bank account number, passwords or any other information by mimicking a website of an enterprise, and the victim is unaware of phishing website. In literature, several approaches have been proposed for detection and filtering phishing attack. However, researchers are still searching for such a solution that can provide better results to secure users from phishing attack. Phishing websites have certain characteristics and patterns and to identify those features can help us to detect phishing. To identify such features is a classification task and can be solved using data mining techniques. In this paper, we are presenting a hybrid model for classification to overcome phishing-sites problem. To evaluate this model, we have used the dataset from UCI repository which contains 30 attributes and 11055 instances. The experimental results showed that our proposed hybrid model outperforms in terms of high accuracy and less error rate.