Learning Log-Based Group Work Support: GLOBE Framework and System Implementations

Group work activities can promote the interpersonal skills of learners. To support the teachers in facilitating such activities, we suggested a learning analytics-enhanced technology framework, Group Learning Orchestration Based on Evidence (GLOBE) with data-driven approaches. We designed and implemented a group formation system using genetic algorithms to form groups using learning log data. Even if there is no existing data, we presented a paradigm of continuous data-driven support for the whole group learning process, incorporating the peer and teacher evaluation results as input to subsequent groupings. Further, utilizing accumulated group learning evidence in such an ecosystem, we aim to explore predictive group formation indicators which can lead to automatic group formation based on teachers' purpose in different contexts for desirable performance in subsequent group learning phases.