Assessing and predicting review helpfulness: Critical review, open challenges and research agenda

The issues of determining what makes a customer review helpful and how to predict review helpfulness have received significant attention in recent years. However, the current literature is highly heterogeneous with contradictory findings.
We critically synthesize extant literature and identify factors responsible for these contradictions which we classify into four dimensions, viz. datasets, features, approach and review helpfulness operationalization, and clarify the prevailing confusion.
We highlight research gaps, and propose novel research directions to bridge them and to advance the state-of-the-art. We provide resources such as synthetic tables, which summarize past research and develop a (future) research roadmap. These resources not only provide a concise overview of current research, but also illustrate its heterogeneity and serve as a valuable basis to ground future research.