Spam Email Assessment Using Machine Learning and Data Mining Approach

Electronic mail is a crucial means of communication, used in both formal and informal ways. Unnecessary communications sent in bulk over the Internet or through any electronic messaging system are known as spam. Some illegal use of spam includes unethical behaviors, phishing, and various other scams. We found that spam emails constitute at least eighty percent of the mail received by the recipients. The key objective of this paper is to provide information for the detection of spam electronic mail. Popular machine learning techniques like K-Nearest Neighbor (KNN), Logistic regression (LR), Cat-Boost, Ada-Boost, Gradient Boosting, Extreme Gradient Boosting (XG-Boost) Light Gradient Boosting Machine (Light GBM), Support Vector Machine (SVM), Decision Tree, Gaussian Naïve Bayes, and Random Forest were implemented on selected features from the data set. Cat-Boost outperformed all the other machine learning algorithms and achieved an accuracy of 0.9699 and an Fl score of 0.97 for spam email classification.