Application of Machine Learning Algorithms for Detection of Vulnerability in Web Applications

The Internet is a world-class network that connects systems and electronic devices. As per the report, 4.66 billion people in the world use the internet for one or other purposes. The internet also provides a wide range of web applications, which provides vast benefits to society and the users. Nowadays, cyberattacks like denial of service (DoS), SQL injections, brute force, and phishing attacks on websites, web applications, and web of things are more common. During the development phase, these security issues need to be addressed efficiently. These internet-based applications, store very critical, valuable, and important information related to user credentials, financial, biometric, payment information, etc. The adversary tries to find vulnerabilities and exploit them to capture the information related to users, and devices. The adversary can also damage the applications and stop them from working. This paper illustrates and analyses the different types of vulnerabilities in detail. Also, this work provides possible solutions to the various attacks. The data for the analysis are collected through the NESSUS tool. The analysis is carried out using Random Forest Classifier, Multinominal Naïve Bayes, Linear SVC, and Logistic Regression. In this work, Linear SVC has 91% accuracy in identifying the type of vulnerability. The algorithm also shows the accuracy of 98% in giving the solutions for the type of attack. © 2022, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.