Detecting Phishing of Short Uniform Resource Locators using classification techniques

Phishing is a technique wherein a hacker clones a legitimate website, email id, graphic design and asks the target to act on the deceptive link for getting sensitive company and personal data. Approximately 78 per cent of cyber-attacks consist of Phishing techniques. There are websites which store the Uniform Resource Locators and domain name of phished sites. However, ample amounts of clones emerge every day. Hence it takes time up to several days to update the websites. There are many techniques in phishing that even experienced technologists cannot detect. This cyber-security threat is not easily identifiable by looking. Association rule mining in Uniform Resource Locators detects phishing. However, with an increase in the shortened Uniform Resource Locators- less information is derived from the uniform resource locator. We use machine learning algorithms to detect phished shorten Uniform Resource Locators. Naive Bayes Theorem, Logistic Regression and Random Forests Algorithm produced an accuracy of 99.4, 96 and 98 per cent respectively.