Fake Product Review Monitoring System Using Machine Learning

As facts and statistics on the web are growing rapidly, online reviews have been a true revolution in the way people purchase products and services. Nowadays, a wide range of e-commerce sites allow the clients to write their reviews about the product that they purchased from that website as these reviews help the brand to understand the customer requirements and the shortcomings in the product. These brands try their best to get good reviews by improving the quality of product from customers as bad reviews affect their business. Often customers need a review of a product before investing in it as it impacts their decision for purchasing it. However, many of the customers are not satisfied after buying the product from a particular website and feel that the reviews are misleading and fake causing blight on all the people even those who actually try to give genuine reviews. On some online platforms, some of the reactions are planted by various frauds which are either hired by an organization or belong to it and try to reduce the product value of competitors by giving negative reviews to their product. And they often provide good reviews to various products designed by their own company. So, we have used sentiment analysis to analyze reviews online and compared the results for two algorithms which are the SVM classifier and Naive Bayes classifier so that the user can determine whether the reviews are genuine or not.