Abstract:
Phishing attacks are one of the emerging serious threats against personal data security. These attacks are often performed by sending out emails that seem to originate from a trusted party. The objective is to deceive the recipient to release sensitive information such as usernames, passwords, banking details, or credentials. The aim of phishing is to steal a user's identity in order to make fraudulent transactions as if the Phisher were the user.
Though a large number of methods have been proposed and implemented for detecting Phishing attacks, a complete solution is missing. A great amount of research is being carried out to solve this problem using rule based method, browser based method and machine learning approaches but still there are insufficient methods that can be used against Phisher's novel attacks which they are able change time to time.
In this Dissertation entitled "A RULE BASED APPROACH FOR DETECTING PHISHING ATTACKS", a solution is proposed for organization wide solution; rule set has been proposed for this system to filter out phishing mails at the perimeter of the organization. A balanced rule set has been used to keep false positive and false negative low.
The proposed strategy has been simulated using C++. The accuracy of the rule set has been compared with existing approach and accuracy of the proposed rule set has been tested on publicly available phishing and non phishing mails. The relation between number of rules and false positive and false negative is also explored
