Credible vs Fake: A Literature Review on Differentiating Online Reviews based on Credibility

Consumer reviews are a valuable source of information and are often the main factor for decisions about purchases. However, not all reviews are credible. Many scholars have attempted to device methods to assess the credibility of online reviews in different disciplines. To date, there is no study that compares research on this topic across these different disciplines. This paper conducts an extensive literature review to provide a comprehensive overview of techniques, classifying them into data-driven and human-driven approaches. Using co-topic analysis, this paper identifies important attributes used to assess the credibility of online reviews. We further classify these attributes into four main categories: review-centric, reviewer-centric, receiver and environmental attributes. Finally, we conclude by presenting and discussing the directions for future research on datasets and attributes that determine the credibility of consumer reviews.