A Systematic Approach for Learner Group Composition Utilizing U-Learning Portfolio

A context-aware ubiquitous learning environment allows applications to acquire diverse learning behaviors of ulearners. These behaviors may usefully enhance learner characteristics analysis which can be utilized to distinguish group learners for further instruction strategy design. It needs a systematical method to analyze ulearner behaviors and utilize learner characteristics for group composition. This paper proposes an effective and systematic learner grouping scheme containing transformation processes from u-portfolios to the proposed Portfolio Grid, creating a learner similarity matrix, and group composition. This study also evaluates intra-group diversity of each resultant heterogeneous group and analyzes learning behavioral patterns acquired from the study experiment. The results indicate that the proposed learner grouping algorithms had positive effects on group composition and interaction between group members for follow-up ubiquitous collaborative learning.