Transformer-Based Automated Fake Review Detection Using Natural Language Processing and Deep Learning

The rapid proliferation of online reviews has led to the rise of fake reviews, which mislead consumers and manipulate market dynamics. This research presents an automated fake review detection system leveraging Deep Learning and Natural Language Processing (NLP) to improve consumer decision-making. The proposed model utilizes Bidirectional Encoder Representations from Transformers (BERT) to analyze textual content and detect deceptive reviews with high accuracy. Benchmark comparisons with five existing techniques—Sentiment Analysis, TF-IDF + SVM, Word2Vec + LSTM, Deep Autoencoder, and CNN + LSTM—demonstrate the superior performance of the BERT-based model. Experimental results on publicly available datasets show that the proposed model achieves an accuracy of 95.6%, precision of 94.8%, recall of 94.2%, F1-score of 94.5%, and an AUC-ROC of 96.2%. These results confirm its effectiveness in identifying both fake and genuine reviews, surpassing conventional methods by up to 20.2%. The model’s high accuracy and efficiency make it suitable for deployment in e-commerce platforms to enhance review credibility and prevent online deception. Future work will focus on improving the model’s robustness through adversarial training, extending its capabilities to multilingual datasets, and optimizing it for real-time large-scale applications. IndexTerms -Fake Review Detection, Deep Learning, NLP, BERT, Transformer Models, Consumer Trust, Sentiment Analysis, Machine Learning, Online Fraud Detection, E-Commerce Security, Text Classification, Opinion Spam, Trustworthy AI, Review Manipulation, Fraudulent Reviews