Detection and Summarization of Honest Reviews Using Text Mining

Abstract

When it comes to purchasing and making business decisions, online reviews have become incredibly valuable. The reviewer can boost brand loyalty while also assisting other customers in understanding their product experience. As internet reviews are becoming more prominent, fraudulent reviews which refer to reviews submitted by authors who are rewarded for creating fake evaluations to influence readers' perceptions, are becoming more common. This research paper aims to create a product review summarizer that generates a summary for amazon product reviews based on non-fake reviews. In the current work, we have compared supervised machine learning algorithms with SVD dimensionality reduction and a text mining approach for the summarization. The labeled amazon review dataset was used for building the model. This paper gives a novel idea of giving text summary of amazon reviews after filtering out the fake ones.