Identification of Fraudulent Reviews

Abstract

This paper introduces a comprehensive system designed to bolster the trustworthiness of product reviews in e-commerce applications. Leveraging logistic regression, the system filters out fake reviews obtained through web scraping, providing users with an authentic product rating. The algorithm analyzes textual features to assign a probability score, effectively distinguishing genuine reviews from deceptive ones. The resultant authentic rating serves as a reliable metric for users navigating the crowded marketplace. In addition to enhancing review authenticity, the system integrates a comparative pricing feature. Multiple e-commerce links are scrutinized to compile and analyze pricing information, enabling users to make well-informed decisions based on both review credibility and cost-effectiveness. The user-friendly interface displays the authentic product rating alongside a graphical representation of the percentage of genuine and fake reviews, empowering consumers to interpret feedback reliability intuitively. The system contributes to e-commerce advancement by addressing the pervasive issue of fake reviews, offering users a sophisticated toolset for assessing product authenticity and making informed purchasing decisions. This research amalgamates machine learning, web scraping, and comparative analysis into a seamless framework, ultimately providing users with a holistic solution for navigating the intricacies of online shopping