DETECTING FAKE REVIEWS IN E-COMMERCE USING MACHINE LEARNING: A CASE STUDY ON AMAZON

Abstract

This paper explores the feasibility of using machine learning algorithms to detect fake reviews on online e-commerce platforms. The recent overreliance of consumers on online reviews to gauge the quality and desirability of e-commerce products has prompted these e-commerce brands to create fake, misleading reviews to shape customer narratives and perceptions about their products. This paper identifies current gaps in the mechanisms employed to filter out fake online reviews and analyzes how Amazon uses emerging methods like behavioral metrics, text analysis, and advanced neural networks to detect fake reviews. Given that a large enough dataset is used, these methods vastly outperform rule-based systems, showing that machine learning methods have significant potential in countering market and narrative distortion in the e-commerce world.