Short Messaging Service (SMS) spam has been known to be the unwanted or unintended messages received on mobile phones. This paper has presented a review of current methods, existing problems, and future research directions on spam classification techniques of mobile SMS spams. The methodology involves collecting a large dataset of SMS messages, both legitimate and spam, to train and evaluate various machine learning algorithms. Feature extraction techniques have been employed to capture relevant information from SMS messages, such as the presence of specific keywords, the length of message, and the sender's identity. The experimental results on the proposed spam filtering system achieves a high level of accuracy with a low false-positive rate, thereby minimizing the chances of legitimate messages being classified as spam. The system effectively detects and blocks a significant portion of spam messages, providing mobile users with a reliable defense against unwanted SMS communications. The findings of this study reveal that machine learning algorithms, particularly ensemble methods like Random Forests, performed well in SMS spam filtering on mobile devices.
Mohamed Abbas Ibrahim (2023) studied this question.