Linguistic Changes across Different User Roles in Online Learning Environment. What do they tell us?

Abstract

In recent years, we have witnessed an increasing interest in online learning environments, particularly in Massive Open Online Courses (MOOCs). However, prevailing studies show that lower percentage of students complete their courses successfully in online learning environment. The vast amount of student data available in MOOC platforms enables us to gain insight into student learning behaviours. In this paper, we explore the idea of ‘student roles’, identifying linguistic change associated with roles that will later help us to understand students’ learning process in MOOCs. As an initial stage of this research, the study aims to categorise student roles (e.g. information seeker, information giver) using discourse analysis, and to further analyse the linguistic change for each student role with time. A multi-class classifier has been built to identify user roles with 82.20% F-measure. Further, our study on linguistic changes demonstrates that distinctive behaviors can be observed across different user roles. Prominent observations include discourse complexity, lexical diversity, level of information embeddedness and lexical frequency profile being high in information giver in comparison to information seeker and other user roles.