EMAIL SPAM FILTERING SYSTEM USING MACHINE LEARNING

Abstract

Email spam filtering has become an essential component of modern communication systems due to the rapid increase in unsolicited and malicious emails. This project presents a machine learningbased email spam filtering system designed to automatically classify emails as spam or legitimate (ham). The system utilizes a labeled dataset of emails and applies preprocessing techniques such as tokenization, stop-word removal, and normalization to prepare textual data for analysis. Feature extraction methods like Bag of Words and TF-IDF are used to convert text into numerical representations. Machine learning algorithms including Naive Bayes, Support Vector Machine, and Logistic Regression are employed to train the model and identify patterns associated with spam content. The trained model is then used to classify incoming emails in real-time, ensuring efficient and accurate filtering. Compared to traditional rulebased systems, the proposed approach adapts to new spam patterns and improves performance over time through continuous learning. The system architecture includes modules for email input, preprocessing, feature extraction, model training, prediction, and output classification. Experimental evaluation demonstrates improved accuracy, precision, and recall in detecting spam emails while minimizing false positives. This system enhances user experience by reducing unwanted emails, protecting users from phishing attacks, and maintaining a clean inbox. The project highlights the effectiveness of machine learning techniques in email security and provides a scalable solution for real-world applications.