Performance Analysis of Machine Learning Algorithm in Spam Email Classification

The proliferation of unsolicited spam emails poses a growing threat to cybersecurity and user productivity. This study investigates the effectiveness of machine learning algorithms which includes Naïve Bayes, Support Vector Machines (SVM), and Random Forest, in accurately detecting spam emails. Using two labeled dataset of spam and legitimate (ham) messages sourced from kaggle, the models were trained and evaluated using standard performance metrics. The Naïve Bayes classifier achieved an accuracy of 94.6%, with a precision of 93.3% and an F1-score of 94.0%. The SVM model demonstrated slightly improved performance with an accuracy of 95.2%, precision of 94.5%, and F1-score of 94.8%. Random Forest outperformed both with the highest accuracy of 96.8%, precision of 95.7%, and F1-score of 96.2%. These findings affirm the viability of machine learning for robust spam detection and highlight the Random Forest algorithm as particularly effective for real-world applications in email filtering systems.