Neural Cognitive Diagnosis

Neural cognitive diagnosis (NCD) aims to automatically assess student knowledge and skills using neural networks, improving upon traditional methods by achieving higher accuracy in identifying individual student proficiencies in specific knowledge concepts. Current research focuses on enhancing the interpretability of these complex models, often employing architectures like Kolmogorov-Arnold networks, and improving student representation by incorporating hierarchical relationships between concepts and student-specific characteristics. These advancements hold significant potential for personalized learning in education and similar applications requiring accurate and insightful assessment of individual capabilities.

Papers