GWpulse: Supporting Learner Modelling and Group Awareness in Online Forum with Sentiment Analysis

Abstract

Collaborative learning in online context has been in educational practice and also gained attention during the emergency remote teaching due to the pandemic. This study follows the Group Learning Orchestration Based on Evidence (GLOBE) framework that considers four phases of technology-supported group work and proposes data-driven services for each phase. In this paper, we present an analysis tool named "GWpulse'' that processes learning log data and shares the analysis results as an attribute of the learner model considered in GLOBE and also presents it in the group work dashboard to facilitate group awareness. In the current version GWpulse considered logs of forum activity generated during group activities in the learning management system. Apart from including that data in the learner model, it also visualizes the activity data for the orchestration, evaluation and reflection phases. GWpulse dashboard displays the basic statistics of their students’ forum activities such as time interval and the number of posts, as well as a novel indicator of "Assistance Needed Level" calculated using sentiment analysis method that classifies textual statements into positive and negative.