ABSTRACT
Nowadays, social media plays a crucial role in our daily life. For many things, we rely on social media. People started believing in something written on the Internet before making any decisions, such as having a look at reviews that are reported on social media for various purposes like buying a product online or booking a hotel room for a vacation, or visiting a place, and others check the reviews to buy products or use various services. Everyone tries to find some uniqueness in multiple activities or products. But there is a lot of controversy going on around these reviews. It has become difficult to see whether a review is genuine or not. To improve their company standards or highlighting their products, they generate a few fake reviews, which attracts the users, and because people start choosing them, they may generate fake reviews about their competent organization products, which leads to a downturn of that particular organization and spoils its reputation. We need a fake review detection system to avoid these fake reviews and go with genuine reviews. This system used some machine learning techniques. The proposed method consists of four algorithms: GBM, XGBoost, LightGBM, and CatBoost. These algorithms help in detecting fake reviews. All these algorithms are compared to each other and display accurate results.
