Consumer preferences for the interface of e-commerce product recommendation system
Journal
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Journal Volume
8527 LNCS
ISBN
9783319072920
Date Issued
2014-06
Author(s)
Abstract
A recommendation system (RS) in a website is increasingly significant for consumer's decision making. A RS includes several important benefits, such as increasing user satisfaction and building user trust. Despite the growing literature that examined the usefulness of a specific attribute of a RS, less is known about which combination of attributes of a RS is preferable and how the combination influences consumer decision making. By using a conjoint analysis, we can further explore the impacts of combination attributes. In a lab experiment, we find that the importance ranking of attributes of a RS for the participants is quite different. Specifically, all the participants consider the attribute, "Explanation for Recommendation", is important. In addition, "Rating" is important for the specific participants. Furthermore, "Comment" seems to be less important to all the participants. Our results have important implications for the design of a RS. © 2014 Springer International Publishing.
Subjects
adaptive interface | conjoint analysis | recommendation system | user interface preferences
SDGs
Type
conference paper