A bidirectional model for analytics-informed assessment and asynchronous learning design

This paper presents a bidirectional model that connects asynchronous learning design and assessment using learning analytics and intelligent optimization. In the forward workflow, interaction data from a Moodle-based LMS are analyzed to identify usage patterns of asynchronous activities and to derive an assessment-oriented ATS (Achievement–Transfer–Stability) profile through fuzzy interpretation. In the backward workflow, a genetic algorithm is used to generate and optimize asynchronous learning paths aligned with a desired ATS target. The results show how learning analytics can support both the analysis of existing course designs and the generation of assessment driven learning activity sequences, contributing to more coherent and stable online learning experiences.