Evaluation of Learning Efficiency in MOOC Teaching Systems Based on Kirkpatrick’s Model and AHP-fuzzy Comprehensive Evaluation Method

Abstract

Conventional single-dimensional scoring methods for Massive Open Online Course (MOOC) teaching systems often fail to capture holistic learning efficiency and suffer from low evaluation accuracy. To address these limitations, this paper proposes a novel learning efficiency evaluation method integrating Kirkpatrick’s four-level model with the Analytic Hierarchy Process (AHP) and fuzzy comprehensive evaluation. First, a multidimensional indicator system is constructed across Reaction, Learning, Behavior, and Results levels. The AHP is then employed to calculate the weight vectors of each indicator, ensuring rational weight distribution through consistency testing. Subsequently, a fuzzy relation matrix is established based on membership degrees derived from expert assessments. By synthesizing the AHP-derived weights with the fuzzy matrix via weighted average operators, the method defuzzifies the results to yield a comprehensive efficiency score. Experimental validation involving 200 learners demonstrates that the proposed method outperforms existing approaches, achieving a Root Mean Squared Error (RMSE) of 0.45, Mean Absolute Error (MAE) of 0.35, and an accuracy rate of 89.5%. The results indicate that this method maintains stable, high-confidence evaluation performance across all levels, effectively handling the fuzziness and uncertainty inherent in learning efficiency assessment, thereby providing a reliable basis for optimizing MOOC teaching systems.