Handling Class Imbalance in SMS Spam Datasets Using Advanced Sampling Techniques

The prevalence of SMS spam poses significant challenges for automated messaging systems, and effective detection is often hindered by the inherent class imbalance in SMS datasets, where legitimate messages vastly outnumber spam messages. This study investigates the impact of advanced sampling techniques on improving classification performance in imbalanced SMS datasets. Techniques such as Synthetic Minority Oversampling Technique (SMOTE), Adaptive Synthetic Sampling (ADASYN), and ensemble-based resampling methods are evaluated for their effectiveness in balancing the dataset and enhancing the predictive accuracy of machine learning classifiers. Experimental results demonstrate that applying these advanced sampling strategies significantly improves spam detection rates while reducing false positives. The findings provide valuable insights for developing robust SMS spam filters and highlight the importance of addressing class imbalance in real-world text classification problems.