Modeling user satisfaction from the extraction of user experience elements in online product reviews

Abstract

With the abundance of product reviews available online, online review data represent invaluable resources for understanding the user experience of various products in their real usage environments. Extant online review studies have considered UX elements mostly related to emotions. We collected 64,772 sentences from 4,380 online reviews of three electronic products, and analyzed the content of the online reviews using LIWC in order to extract various UX elements going beyond emotions. The study results show that UX elements extracted from online reviews had significant effects on user satisfaction. In addition to the emotional factors (hedonic, user burden), the results show that expectation confirmation and pragmatic factors play significant roles in determining user satisfaction. Copyright © 2017 by the Association for Computing Machinery, Inc. (ACM).