Automated Detection of Deceptive Online Product Reviews Using Supervised Learning Techniques

This paper presents a machine learning–based system for detecting fake product reviews using Natural Language Processing (NLP) techniques. With the rapid growth of e-commerce, online reviews significantly influence consumer purchasing behavior, but the rise of deceptive or manipulated reviews has undermined their reliability. The proposed model utilizes text preprocessing methods such as tokenization, stop-word removal, and stemming, followed by feature extraction using TF-IDF and CountVectorizer. Multiple supervised learning algorithms, including Logistic Regression, Random Forest, Decision Tree, Naïve Bayes, and K-Nearest Neighbors (KNN), were implemented to classify reviews as genuine or fake. Experimental results show that the Support Vector Machine (SVM) achieved the highest accuracy of approximately 88.5%, outperforming other models. Analysis of feature importance and confusion matrices revealed that linguistic and frequency-based attributes play a key role in deception detection. The developed system also includes a real-time review classification module, demonstrating its potential for application in Deceptive Review Detection, review moderation, and consumer trust enhancement.