An Efficient Ensemble approach for Fake Reviews Detection

Abstract

People prefer to buy items and services online to save time in the current age of growing e-commerce. these internet purchases are heavily impacted by the reviews or opinions of people who have already purchased them. customers provide comments to businesses on how to improve product quality, develop, and monitor business strategies in order to boost sales and profits. customers may also use these comments to choose the proper items with less effort and time spent. giving fake review is the practice of fraudulent people who wants to promote or degrade products or services for illegitimate monetary gain. in this research paper, we present an ensemble machine-learning model to identify whether a review is fraudulent or authentic. to achieve this objective, amazon reviews dataset is used. the proposed ensemble model outperformed as compared to other individual classifiers. random forest provides 99% accuracy which is better than other algorithms.