An NLP-Based Approach to Modelling of Consumers Perceived Quality and Intention to Recommend in Paid Massive Open Online Courses (Moocs): Mediating Roles of Consumption Emotions

Abstract

Paid massive open online courses (MOOCs), which dominants most commercial MOOC platforms, contribute largely to the sustainable development of quality distance education in supporting personal development, professional advancement, and even economic opportunity. However, little attention has been paid to the post-consumption behavioral intentions of those who paid MOOCs. To study this, we utilize a novel mixed-method approach based on natural language processing (NLP) techniques drawing upon 14,585 online reviews from a well-known global commercial MOOC platform, Udemy.com. Perceived quality factors, post-consumption emotions and intention to recommend have been identified by topic modelling technique. The relationships among those factors are examined from a cognition-affect-behavior perspective using partial least squares structural equation modeling (PLS-SEM). The result suggests that positive consumption emotions positively mediate the relationships between course quality, instructor quality and platform quality and intention to recommend, while negative consumption emotions negatively mediate the relationships. Interestingly, course quality displays a positive effect on negative emotions while two other quality factors show negative effects on negative emotions. These findings offer some insights into the emergency paid knowledge market and suggestions for instructional design of online courses.