Abstract:Cognitive diagnosis is a fundamental issue of intelligent education platforms, whose goal is to reveal the mastery of students on knowledge concepts. Recently, certain efforts have been made to improve the diagnosis precision, by designing deep neural networks-based diagnostic functions or incorporating more rich context features to enhance the representation of students and exercises. However, how to interpretably infer the student's mastery over non-interactive knowledge concepts (i.e., knowledge concepts not related to his/her exercising records) still remains challenging, especially when not giving relations between knowledge concepts. To this end, we propose a Knowledge-Sensed Cognitive Diagnosis (KSCD) framework, aiming at learning intrinsic relations among knowledge concepts from student response logs and incorporating them for inferring students' mastery over all knowledge concepts in an end-to-end manner. Specifically, we firstly project students, exercises and knowledge concepts into embedding representation matrices, where the intrinsic relations among knowledge concepts are reflected in the knowledge embedding representation matrix. Then, the knowledge-sensed student knowledge mastery vector and exercise factor vectors are obtained by the multiply product of their embedding representations and the knowledge embedding representation matrix, which make the student's mastery of non-interactive knowledge concepts be interpretably inferred. Finally, we can utilize classical student-exercise interaction functions to predict student's exercising performance and jointly train the model. In additional, we also design a new function to better model the student-exercise interactions. Extensive experimental results on two real-world datasets clearly show the significant performance gain of our KSCD framework, especially in predicting students' mastery over non-interactive knowledge concepts, by comparing to state-of-the-art cognitive diagnosis models (CDMs).

Cognitive Diagnosis with Explicit Student Vector Estimation and Unsupervised Question Matrix Learning

A Semi-supervised Learning Method for Q-Matrix Specification Under the DINA and DINO Model With Independent Structure

Cognitive diagnostic assessment: A Q-matrix constraint-based neural network method

Novel Use of Self-organizing Map for Q-matrix Calibration in Cognitive Diagnosis Assessment

QCCDM: A Q-Augmented Causal Cognitive Diagnosis Model for Student Learning

Interpretable Cognitive Diagnosis with Neural Network

Interpretable Cognitive Diagnosis with Neural Network for Intelligent Educational Systems.

Neural Cognitive Diagnosis for Intelligent Education Systems

Fuzzy Cognitive Diagnosis for Modelling Examinee Performance

Regularized Bayesian algorithms for Q‐matrix inference based on saturated cognitive diagnosis modelling

Using country-specific Q-matrices for cognitive diagnostic assessments with international large-scale data

Knowledge-Sensed Cognitive Diagnosis for Intelligent Education Platforms

Towards a New Generation of Cognitive Diagnosis

Modeling question difficulty for unbiased cognitive diagnosis: A causal perspective

Cognitive Diagnosis Focusing on Knowledge Concepts

Cognitive Diagnostic Model Made More Practical by Genetic Algorithm.

Modeling Context-aware Features for Cognitive Diagnosis in Student Learning

Icd: A New Interpretable Cognitive Diagnosis Model for Intelligent Tutor Systems

Learning Evidential Cognitive Diagnosis Networks Robust to Response Bias.

Cognitive Diagnosis of Students' Test Performance Based on Probability Inference

An exploratory Q-matrix estimation method based on sparse non-negative matrix factorization