Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge

Miao Zhang, Ziming Wang, Runtian Xing, Kui Xiao, Zhifei Li, Yan Zhang, Chang Tang

TL;DR

Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs.

Abstract

Cognitive diagnosis is a fundamental and critical task in learning assessment, which aims to infer students' proficiency on knowledge concepts from their response logs. Current works assume each knowledge concept will certainly be tested and covered by multiple exercises. However, whether online or offline courses, it's hardly feasible to completely cover all knowledge concepts in several exercises. Restricted tests lead to undiscovered knowledge deficits, especially untested knowledge concepts(UKCs). In this paper, we propose a novel framework for Cognitive Diagnosis called Disentangling Heterogeneous Knowledge Cognitive Diagnosis(DisKCD) on untested knowledge. Specifically, we leverage course grades, exercise questions, and learning resources to learn the potential representations of students, exercises, and knowledge concepts. In particular, knowledge concepts are disentangled into tested and untested based on the limiting actual exercises. We construct a heterogeneous relation graph network via students, exercises, tested knowledge concepts(TKCs), and UKCs. Then, through a hierarchical heterogeneous message-passing mechanism, the fine-grained relations are incorporated into the embeddings of the entities. Finally, the embeddings will be applied to multiple existing cognitive diagnosis models to infer students' proficiency on UKCs. Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs. Our code is available at https://github.com/Hubuers/DisKCD.

Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge

TL;DR

Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs.

Abstract

Cognitive diagnosis is a fundamental and critical task in learning assessment, which aims to infer students' proficiency on knowledge concepts from their response logs. Current works assume each knowledge concept will certainly be tested and covered by multiple exercises. However, whether online or offline courses, it's hardly feasible to completely cover all knowledge concepts in several exercises. Restricted tests lead to undiscovered knowledge deficits, especially untested knowledge concepts(UKCs). In this paper, we propose a novel framework for Cognitive Diagnosis called Disentangling Heterogeneous Knowledge Cognitive Diagnosis(DisKCD) on untested knowledge. Specifically, we leverage course grades, exercise questions, and learning resources to learn the potential representations of students, exercises, and knowledge concepts. In particular, knowledge concepts are disentangled into tested and untested based on the limiting actual exercises. We construct a heterogeneous relation graph network via students, exercises, tested knowledge concepts(TKCs), and UKCs. Then, through a hierarchical heterogeneous message-passing mechanism, the fine-grained relations are incorporated into the embeddings of the entities. Finally, the embeddings will be applied to multiple existing cognitive diagnosis models to infer students' proficiency on UKCs. Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs. Our code is available at https://github.com/Hubuers/DisKCD.
Paper Structure (25 sections, 17 equations, 4 figures, 4 tables)

This paper contains 25 sections, 17 equations, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Cognitive Modeling
  4. Knowledge Coverage Problem
  5. Preliminaries
  6. Cognitive Diagnosis Model
  7. Problem Setup
  8. Heterogeneous Relation Graphs
  9. Methodology
  10. Embedding Layer
  11. Student Embedding
  12. Exercise embedding
  13. Concept embedding
  14. Heterogeneous Relation-aware Layer
  15. Heterogeneous Relation Attention.
  16. ...and 10 more sections

Figures (4)

  • Figure 1: Examples of cognitive diagnosis. Students' mastery of tested knowledge components is often emphasized, while their grasp of untested knowledge components is usually overlooked.
  • Figure 2: Overview of DisKCD. (1) The left side shows the disentangling of knowledge concepts and the original embedding of students and exercises. (2) The middle part shows the details of the heterogeneous relation-aware layer where fine-grained relations are merged into entity embeddings. (3) The right part shows the diagnostic process of DisKCD framework.
  • Figure 3: Overview of AUC, ACC, RESM, and interpretability performance of DisKCD and the baseline across five datasets.
  • Figure 4: Comparison of diagnostic results for two students under the DisK-RCD and RCD models.