DEVELOPMENT OF AN SMS SPAM CLASSIFIER SYSTEM USING MACHINE LEARNING ALGORITHM

There is rapid increase in unsolicited text messages which has raised concerns about privacy, fraud, and communication security. This project presents SMS Guard, an intelligent system developed to automatically detect and classify SMS messages as either spam or ham (legitimate) using machine learning techniques. The system leverages a labeled SMS dataset that underwent text preprocessing, including tokenization, stop-word removal, and vectorization. Several machine learning models including Naïve Bayes, Logistic Regression, Random Forest, and Support Vector Machines were trained and evaluated. A consensus (ensemble) of these models was adopted to generate the final classification, ensuring higher accuracy and reliability. The developed web-based application features a simple, user-friendly interface where users can input messages for analysis and receive real-time feedback. Results include a color-coded verdict, confidence score, and an explanation view showing the words that most influenced the prediction. Additional features such as user authentication, profile management, and a prediction history dashboard were implemented to enhance personalization and usability. The system achieved strong performance and demonstrated that machine learning can effectively mitigate the growing problem of SMS spam. Its explainable design and adaptability make it suitable for further integration into mobile platforms and other communication systems to improve user safety and trust in digital communication.