A Phishing Detection Framework for Websites Based on Data Mining

Abstract

Detecting any Phishing site is extremely an intricate and dynamic issue including numerous variables and criteria. Due to the ambiguities associated with phishing location, fluffy information mining procedures can be a viable instrument in detecting phishing websites. In this paper, we propose a strategy which consolidates fluffy rationale alongside information digging algorithms for detecting phishing websites. Here, we characterize 3 diverse phishing types and 6 unique criteria for detecting phishing websites with a layer structure. We have utilized the SVM Data Mining algorithm for classification. Besides, this we also compare the efficiency of SVM in terms of time and space complexity with Naive Bayes algorithm, the framework proactively disposes of the Phishing site or Phishing page by sending a notice to the System Administrator of the host server that it is hosting a Phishing site which may result in the evacuation of the site. Moreover, in the wake of ordering the Phishing email, the framework recovers the location, IP address and contact data of the host server.