A YouTube Fraud Comments Detection Scheme Using Cascaded Ensemble Machine Learning Model

YouTube has become a major platform for content sharing, but it also faces significant challenges due to fraudulent comments, including spam, scams, and misleading information. Traditional rule-based and standalone machine learning models often fail to detect fraud comments with high accuracy due to evolving spam tactics and diverse comment structures. This paper proposes a cascaded ensemble machine learning model to improve fraud comment detection on YouTube. The proposed system combines multiple classifiers, including Random Forest, Support Vector Machine (SVM), and Deep Neural Networks (DNN), in a cascaded framework to enhance detection accuracy while minimizing false positives. Experimental evaluations on real-world YouTube datasets demonstrate the effectiveness of our approach in identifying fraudulent comments with improved precision and recall, contributing to a safer online community.