SPAM MESSAGE DETECTION OVER SOCIAL MEDIA: A SUPERVISED SAMPLING APPROACH FOR THE SOCIAL WEB OF THINGS

The increasing use of social media has led to a surge in spam messages, including fake advertisements, phishing links, and misinformation. Traditional spam detection methods struggle with evolving spam patterns and imbalanced datasets, where spam messages constitute only a small fraction of total messages. This paper proposes a supervised sampling approach for spam detection in the Social Web of Things (SWoT), leveraging machine learning and natural language processing (NLP) techniques. The system uses Synthetic Minority Oversampling Technique (SMOTE) and cost-sensitive learning to handle class imbalance, improving the classification accuracy of spam detection models. Experimental results on real-world social media datasets demonstrate that the proposed approach enhances spam detection performance, reducing false positives and improving precision.