Informational influence and its forecasting in e-commerce

This chapter presents the effects of social influence in electronic-commerce and the significance of information conveyed by websites. It discusses how other consumers affect the decision making of a buyer by posting information and demonstrates how a machine-learning model can forecast future ratings and reviews. The chapter examines three e-commerce mechanisms in relation to informational influence: online auctions, numerical ratings and textual reviews. Attention is drawn to their ability to facilitate bidirectional communication between consumers. The chapter compares characteristics of these mechanisms and describes how social influence, informational influence, and the Elaboration Likelihood Model (ELM) operate in affecting consumers’ decisions. The key results show that like-minded people’s choices can be learned and modeled. Based on this learned model together with very few initial ratings, the wisdom of crowds can be predicted, demonstrating that abundance of information is not always required. Furthermore, results from the study reveal that the prediction of the sentiment of reviews is more accurate than the prediction of numerical rating. Understanding how consumers respond to information presented to them allows websites to improve their interfaces and user interaction mechanisms.