Review-Based Screening Interface for Improving Users' Decision Process in E-commerce

摘要

Due to the important role of product reviews in e-commerce, some systems have employed different approaches to present reviews on product detail page or comparison page. However, little work has investigated how to present reviews on screening interface for facilitating users to screen out interesting alternatives from a set of options. In this paper, we have developed a novel review-based screening interface in terms of users' behaviors. Concretely, there are two innovations in the interface design: (1) it enables users to eliminate items by sentiment attributes (i.e., the attributes extracted from reviews), and (2) it emphasizes on visualizing the value distribution of each attribute and tradeoffs among attributes. The results of a user study show that our review-based screening interface achieves significantly more positive assessments than traditional screening interface, in terms of users' perceived decision accuracy, cognitive effort, pleasantness and intention to use.