
A global knowledge ecosystem of students, professionals & industry-leaders to share, create, and mentor experiences
contact@podareduspace.org
+91 98202 27795

A Machine Learning Approach for Detecting Suspicious Banking URLs by Harshil Dungarpuria
Digital banking has experienced rapid growth in recent years and has made access to financial services more convenient. However, it has come with a cost of increased cybersecurity problems such as phishing attacks. It is an attack where the user receives a legitimate-looking banking URL that directs the user to a fake website designed to steal sensitive financial and personal details. As the users may not be aware of the official website address, they find it difficult to differentiate between a malicious and a legitimate banking URL. This creates the need for a system that can easily distinguish between banking URLs before users click on them.
A Machine Learning Approach for Detecting Suspicious Banking URLs
- by Harshil Dungarpuria
Abstract
Digital banking has experienced rapid growth in recent years and has made access tofinancial services more convenient. However, it has come with a cost of increased cybersecurity problems such as phishing attacks. It is an attack where the user receives a legitimate-looking banking URL that directs the user to a fake website designed to steal sensitive financial and personal details. As the users may not be aware of the official website address, they find it difficult to differentiate between a malicious and a legitimate banking URL. This creates the need for a system that can easily distinguish between banking URLs before users click on them.
This research proposes a machine learning approach for detecting suspicious banking URLsbefore the users unknowingly access the fraudulent websites. This study’s proposed system analyses the entered URL using features extracted directly from the URL itself such as URL length, the presence of HTTPS, suspicious keywords, and special characters. Based on these features, the system decides whether the URL is malicious or legitimate, providing a protective layer against phishing attacks.
For this study, a labelled dataset containing legitimate and fraudulent banking URLs were used to train the proposed model. Relevant features, such as URL length, the presence of HTTPS, hyphens, and other URL features, were extracted from each URL before being used for machine learning. The Random Forest classifier is used and trained to understand patterns related with legitimate and phishing URLs and distinguish them as legitimate or malicious.
Introduction
Digital banking is the change of traditional banking services to digital format, enabling the customers to manage their finances, perform transactions, and access various other services through digital channels. This all can be done through online banking, mobile banking, and mobile applications. It has become an important part of modern financial services allowing users to perform activities through the internet without visiting any physical bank. With the increasing use of smartphones, internet, and digital payments, online banking has become faster and more convenient. Users can transfer money, pay bills, check the account, etc. However, as the dependence on digital banking continues to grow, safeguarding users and their financial details has also become a challenging task.
To Read The Full Research Paper, Click Below -
