Malicious Transaction URL Detection Using Logistic Regression

The onset of COVID-19 has shifted people to the virtual world more to avoid social interaction. In fact, the trading systems have shifted to E-Commerce platforms which include virtual transactions as well. Besides being a boon, virtual transactions have also brought in several cybercrimes. Several fraudulent sites are created for distracting the users from authentic sites and lead them to the unsafe ones where users lose their personal data or even their wealth at worst cases. This paper is to illustrate a machine learning solution for putting a check on these kinds of activities. Different Python tools have been used to preprocess the dataset that contains different URLs which is classified into safe unsafe categories depending on the presence of SSL certificate or encryption of the URL which is further fed into a model that classifies URLs into safe and unsafe categories using logistic regression. Other than earlier approaches, our proposed model not only checks simple URLs but also check whether URLs are safe for transaction.