A Cognitive Diagnostic Model that Integrates the Relationship between Knowledge Concepts and the Significance of Knowledge Points

Cognitive diagnosis is crucial for evaluating learners<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">’</sup> cognitive traits and skills. Current methods often rely on linear or logical functions, limiting their ability to handle complex relationships and requiring manual labeling. This paper proposes the CK-CD model, which addresses these limitations by integrating knowledge concept enhancement and knowledge point importance. The model employs self-attention mechanisms to uncover hidden relationships between concepts and uses attention mechanisms to explore the importance of knowledge points within each item. This comprehensive approach improves diagnostic accuracy and provides valuable insights into learners’ cognitive processes. Experimental results on a real-world dataset demonstrate the effectiveness of the CK-CD model in predicting student performance, outperforming traditional and neural network-based models