ABSTRACT

E-Commerce is one of the most flourishing businesses in today’s world. A large part of the population, especially in urban areas, is switching towards e-commerce websites to fulfill all of their shopping requirements, whether groceries, electrical appliances, clothing, etc. In an online purchase, product review is considered a significant factor in deciding the right choice of product. Therefore, e-commerce businesses are primarily dependent on product reviews. Due to the lack of authenticity of the reviewer information, while posting a review of any product or service online, the presence of fake reviews is increasing day by day. The presence of these fake reviews of various products or services impacts the customers and the sellers. The customers might choose the wrong brand of a product or service, while the sellers might face low sales of their high-quality products because of these fake reviews. This paper used different machine learning approaches to detect fake reviews of services on e-commerce sites. We have further categorized the fake reviews into positive and negative based on the reviewer’s rating.

Keywords: - Logistic regression, Support Vector Machine, Decision Tree, Random Forest, Neural Networks