Email Spam Detection Using Naive Bayes and Random Forest Classifiers

Abstract

AbstractEmail is used in business and education and almost everywhere for communications. Email is categorized into many categories based on its content like primary, social, promotional, and spam. So spams are also a sub-category under all the categories. Spam may be in the form of text messages,web messages, images and others also. In this paper, we are discussing Email spam. Email spam, sometimes termed junk emails or undesired emails that consumes computing resources, users time and information. These emails are also involved in many cyber crimes like phishing, vishing, identity stolen, data stolen and more. Spam detection and filtration are major issues for email providers and users. Email filtering is a key tool for detecting and combating spam. In our research work, Machine learning algorithms Naive Bayes and random forest are utilized to achieve the objective of spam detection. To evaluate performance, we use accuracy, precision, recall, and f score metrics and discussed the results. We found that these classifiers performed well. In the future, we can utilize deep learning to achieve better results in both text and email categories.KeywordsCyber securityCyber crimeSPAMNaive BayesRandom Forest