Understanding Learner Behavior Using Information Theory on Learning Analytics and Knowledge

Abstract

It has been challenging for traditional Learning Analytics (LA) to accurately deal with learners’ knowledge acquisition due to its data-driven nature. As a solution to this problem, this paper introduces Information Theory on Learning Analytics and Knowledge (ITLAK), a theoretical framework that quantifies the information value of learning behaviors. By defining the entropy of learning behavior and modeling learning as an information-theoretic process, ITLAK provides principled measures of knowledge engagement and entropy based on learner interactions with knowledge elements. The framework provides various applications, including personalized recommendations, while offering interpretable indicators of learning diversity and progress. It also supports the diagnosis of conceptual imbalance and contributes to theory-informed learning support. ITLAK advances LA by shifting focus from surface-level activity analysis to epistemically grounded models of knowledge acquisition. ITLAK may help bridge theory and practice toward more effective and interpretable learning support through future developments and research, such as empirical implementation and temporal simulation.