Visualizing Knowledge-Based Learning Logs to Support Multi-Perspective K-12 Math Instruction

Abstract

Conventional learning analytics (LA) systems often rely on test scores and behavioral logs, providing limited support for understanding learners from multiple pedagogical perspectives, which are essential for personalized instruction. Recognizing the need for more accessible and beneficial LA tools for K-12 teachers, we developed a visualization dashboard based on the Open Knowledge and Learner Model (OKLM) that integrates diverse learning data into a unified structure to enable flexible, multi-perspective interpretation of students’ learning states. Through interviews with three junior high school mathematics teachers, we explored how the dashboard supports teachers in interpreting learners’ states from various pedagogical perspectives. The results suggest that perspectives such as self-awareness and relative positioning were especially helpful for understanding students, and that the system allows flexible interpretation aligned with instructional goals. These findings help realize the potential of OKLM to offer richer insights for personalized instruction in real-world educational contexts.