Date of Award

Fall 2025

Project Type

Thesis

Program or Major

Computer Science

Degree Name

Master of Science

First Advisor

Dongpeng Xu

Second Advisor

Laura Dietz

Third Advisor

Sean Moore

Abstract

Phishing domains are a common and effective attack vector that malicious actors use to fool users into believing that they are going to a trusted website in the hopes of compromising users. These attacks can also be referred to as phishing attacks because the domain in which malicious attackers act has a hook to trick users into going to the domain that users believe is legitimate. These malicious actors mislead users into believing they are visiting the intended site, when in reality they have been led to a website hosted by the actors. The purpose of this paper is to introduce Phishing Hunger as a method to detect these phishing domains in real time. The method takes a potential malicious domain name and outputs whether the domain name is a phish or not with the inverse phish or legitimate domain. This determination is made by finding the closest known legitimate domain based on the structure and makeup of the domain. By providing the decision of whether a domain is a phish, the context of how that decision was made, and what the phishing domain is trying to impersonate. Using two datasets of domains labeled as phishing or not, Phishing Hunter was able to achieve accuracy scores as high as 91\% when a match was found. There were many domains where a determination could not be made, showing that there is room for improvement for the system. Phishing Hunter can identify phishing domains and provide what that domain is attempting to impersonate with context.

Download

Share

COinS