Predicting examinee performance based on a fuzzy cloud cognitive diagnosis framework in e-learning environment.

Area of Science:

• Educational Technology
• Artificial Intelligence in Education
• Cognitive Science

Background:

• Cognitive Diagnosis Models (CDMs) assess learner skills for performance prediction and personalized e-learning.
• Existing CDMs face limitations with complex skills, uncertain proficiency, and large datasets, impacting measurement and efficiency.
• Accurate prediction of learner performance is crucial for effective e-learning strategies.

Purpose of the Study:

• To propose a novel fuzzy cloud cognitive diagnosis framework (FC-CDF) for predicting examinee performance in e-learning environments.
• To address the limitations of existing CDMs in measuring skill proficiency and diagnostic efficiency.
• To enhance decision-making support for personalized e-learning instruction through accurate performance prediction.

Main Methods:

• Utilized normal cloud models (NCMs) to measure expectation, variation, and frequency of learner skill proficiency.
• Transformed NCMs into interval fuzzy numbers to represent the uncertainty of skill proficiency.
• Integrated educational psychology hypotheses with parameter estimation to determine skill levels and item slip/guess factors.

Main Results:

• The proposed FC-CDF accurately predicts examinee performance in e-learning.
• The approach significantly reduces execution time compared to existing methods.
• Demonstrated improved measurement mechanisms and diagnostic efficiency for cognitive diagnosis.

Conclusions:

• The FC-CDF offers a robust framework for cognitive diagnosis and performance prediction in e-learning.
• This method effectively handles skill complexity and learner proficiency uncertainty.
• The findings support enhanced personalized instruction and decision-making in digital learning environments.