SMS SPAM DETECTION USING MACHINE LEARNING

The rapid development of mobile technology has brought with it the challenge of dealing with SMS spam, which has become a major concern for users’ privacy and telecommunication systems within networks. Apart from causing inconveniences, SMS spam can also lead to phishing, financial fraud, and even the spread of malware. This paper focuses on recent studies on SMS spam detection that utilized Machine Learning (ML) and Deep Learning (DL) technologies. Focus is given to model selection, dataset compilation, preprocessing steps, evaluation benchmarks, and explainability for performance assessment. The most accurate models are the state-of-the-art hybrid deep and transformer-based models because of their flexibility in capturing complex patterns within the text, although traditional ML approaches are still applicable for resource-constrained, lightweight deployments.