Phishing Urls Detection System Using Lexical Feature Analysis

Abstract

Phishing website aims to steal your account password or any other confidential data by making you
to believing that you’re on a legitimate website. You can even land on a phishing website by clicking
on suspicious URL. These URLs are called malicious URLs. The inability of the end user system to
detect and remove the malicious URLs can put the legitimate user in vulnerable condition. There are
number of solutions to identify phishing websites have been proposed. These methods fetch website
content which results in undesired side effects. Also there are solutions like black-listing. Blacklisting contains the list of all URLs which are previously known to be malicious. But the list is not upto-date and also it is a time consuming task. To overcome this problem, a classification model is
developed to identify these dynamically changing URLs. In this paper we present a model which is
built using lexical feature analysis that not only takes care about syntactical nature of the URL, but
also the type of attack which the URL is undergone. The type of attack is determined based on the
keywords present in the URL. Moreover, the frequency of a particular type of attack is specified, so
that we can identify which type of attack that we are highly facing by the use of generic framework.
And hence user can take the preventive measures by providing more security against frequent
attacks. For better analysing, we presented a graphical visualization on the frequency of attacks.