Abstract

An efficient SMS spam detection system is developed using the Multinomial Naive Bayes (MNB) algorithm. It employs a labeled dataset and extracts features with the term frequency-inverse document frequency (TF- IDF) method. The MNB algorithm classifies messages by modeling term probability distributions. Parameter tuning and pre-processing techniques like text normalization and stop-word removal enhance feature quality. Experimental results show high accuracy, precision, recall, and F1-score, making MNB suitable for real-time applications. The system provides a practical solution for SMS spam detection and enhances mobile communication security.