Detecting Fake Reviews on E-commerce Platforms Using Machine Learning

Currently, reviews on different e-commerce websites are fake and mostly posted by the company itself to increase the reach or sale of the product itself. These reviews are mostly generated by different bots or written manually. This study provides a novel framework that combines machine learning (ML) and natural language processing (NLP) methods to identify fraudulent reviews on e-commerce platforms. The system is subjected to extensive data preparation and feature extraction, utilising cutting-edge natural language processing techniques to identify subtle linguistic patterns and emotions. By utilising supervised and unsupervised machine learning methods, together with ensemble learning for enhanced precision, the model successfully differentiates between authentic and fraudulent reviews. Experiments conducted on real-world datasets show that the system performs robustly, avoiding false positives, while maintaining high detection rates. This study makes a substantial contribution to improving the integrity of e-commerce platforms by offering workable solutions to prevent fraud and protect customer confidence.