Abstract
Phishing is one of the cybercrime methods which seems to be increasing steadily targeting unsuspecting users. Nowadays, the impact of phishing extended to include organizations that provide services via the Internet. Such organizations become more susceptible to lose their reputation and their competitive edges because of phishing. Classifying a website into a phishing or legitimate depends mainly on the status of a set of significant attributes that exist in the website. Various solutions have been developed to mitigate phishing attacks. Yet, there is no solution that was able to solve the problem completely. One of the encouraging methods that can be utilized is the data mining. Specifically, the "induction of classification rules". In this article, a novel association classification algorithm is suggested and applied to the well-known phishing websites dataset in the UCI repository. The experimental results were promising with respect to several evaluation criteria that are commonly utilised in evaluating classification data mining domains.