The rapid growth of e-commerce platforms has led to an increasing reliance on customer reviews for purchasing decisions. However, the prevalence of fake reviews poses a significant challenge that undermines consumer trust and product ratings. This study addresses the issue of fake comments in Amazon product reviews by employing a comparative analysis of machine learning models like Support Vector Machines (SVM), Classification and Regression Trees (CART) and Multilayer Perceptron (MLP). Amazon reviews’ dataset was preprocessed through text normalization, feature extraction and also feature selection techniques (TF-IDF). These results demonstrate that SVM achieves the highest accuracy (90%) and specificity (92%), outperforming CART and MLP in detecting fake reviews.