A review of research on detection of fake commodity reviews

Abstract

With the development of the Internet era, more and more consumers are shopping online. Product reviews can influence consumers' consumption decisions. Fake review detection is a valuable research. From the perspective of research data content, it can be divided into two categories: Review content based and review metadata; from the perspective of fake review detection method, it can be divided into four categories: Rule-based, graph-based model, machine-learning based and deep learning based. Common data sets are also introduced.