Strengthening Cybersecurity: A Comparative Study of KNN and Random Forest for Spam Detection

At present, email has become a necessary communication medium for exchanging messages and is considered an important part of business, commerce, government, education, entertainment, and other fields in various countries. As it is known that everything has its pros and cons, in spite of benefitting society everybody have to face its drawbacks, which include email spamming. Email spam or junk emails are unwanted emails which are annoying as well as dangerous, containing links trying to corrupt the computer system, stealing your bank details, and stealing your identity with the help of Botnet or real humans, fraudulent access to the information through data breaches. Spam filtering algorithms detect undesired, infected mail and block these messages from reaching to user’s inboxes and keeping their email servers safe from getting overloaded. These are adaptable and provide sustainability to all the spam detected and provide security to the emails. It is important to make the network and system to be free from spammers, malicious links, and viruses. In this research, it has been demonstrated that the K-Nearest Neighbor algorithm by storing the training data and the class labels and is used for classification as well as regression and Random Forest containing more than one decision tree on different subsets improves the accuracy and achieve high accuracy rates, often above 95% by classifying email into spam and ham.