Various Machine Learning Models for Classifying Trash Email

Abstract

Unsolicited or trash emails have become a major issue in modern digital communication, creating significant inconvenience and potential risks for users. These emails not only waste time and resources but may also contain malicious links, phishing attempts, or harmful attachments. As a result, spam detection has become an important area of research, and Artificial Intelligence (AI) models are increasingly applied to improve filtering accuracy. In this project, several machine learning models were employed to classify the spam and legitimate emails. The dataset was preprocessed, and the classification results were not only analyzed in text form but also visualized through charts to provide a clearer comparison of performance. Different algorithms, including traditional classifiers and advanced models, were tested and evaluated. The results demonstrate that AI-based approaches can achieve high accuracy and efficiency in identifying spam emails, highlighting their effectiveness as reliable tools for addressing the persistent problem of unwanted email communication.