Spam mail detection various machine learning methods and their comparisons
Harendra Singh Negi, Aditya Bhatt, Vandana Rawat · Algorithms · 2024
The application of machine learning techniques for spam mail detection is explored in this chapter. Utilizing both the CountVectorizer and TF-IDF vectorizer techniques, five algorithms were created: naive Bayes, decision tree, random forest (RF), support vector machine (SVM), and XGBoost. Performance metrics such as AUCROC, precision-recall curve, F1 score, recall, accuracy, and precision were utilized to evaluate each method. With an accuracy of 96.67%, RF outperformed the other algorithms while using CountVectorizer. SVM and RF were found to be the top-performing algorithms by using TF-IDF vectorizer, with accuracy ratings of 97.33% and 96.50%, respectively. The results imply that these algorithms are effective in detecting spam mail. The study’s conclusions on the effectiveness of several machine learning algorithms for spam mail identification might be useful to researchers in the domains of natural language processing and machine learning.