RCD: Relation Map Driven Cognitive Diagnosis for Intelligent Education Systems

This is our implementation for the paper:

Weibo Gao, Qi Liu*, Zhenya Huang, Yu Yin, Haoyang Bi, Mu Chun Wang, Jianhui Ma, Shijin Wang, and Yu Su. RCD: Relation Map Driven Cognitive Diagnosis for Intelligent Education Systems[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2021: 501-510.

Please cite our SIGIR'21 paper if you use our codes. Thanks!

Author: Weibo Gao (http://home.ustc.edu.cn/~weibogao/)

Email: [email protected]

Environment Settings

We use Torch and DGL as the backend.

• Torch version: '1.7.1'
• DGL version: '0.6.1'

Example to run the codes.

The instruction of commands and take Junyi dataset as an example (We will provide ASSIST dataset as soon as possible).

Go to the code directory:

cd RCD/RCD

Create two folders '/model' and '/result'.

Build exercise-concept correlation local map:

python build_k_e_graph.py

Build student-exercise interaction local map:

python build_u_e_graph.py

Train and test RCD model:

python main.py

Note: In subsection 4.3 (i.e., Extendable Diagnosis Layer) of the paper, Q_{e} in original MIRT represents exercise discrimination. We use a concept-related vector instead of discrimination as an implementation in the paper. RCD can be extended to the many forms of cognitive diagnosis.

Dataset(RCD/data)

junyi

log_data.json:

• Student exercising records.
• Source: https://github.com/bigdata-ustc/EduData

train_set.json

• Data file for training.

test_set.json

• Data file for testing.

graph/K_Directed.txt

• Prerequisite relation from concept dependency local map.
• Each line is a prerequisite relation from concept dependency local map: precursor_concept_ID\t succeed_concept_ID.

graph/K_Undirected.txt

• Similarity relation from concept dependency local map.
• Each line is a similarity relation from concept dependency local map: concept_ID\t similar_concept_ID.

Note: Exercise-concept correlation local map and student-exercise interaction local map can be constructed by running build_k_e_graph.py and build_u_e_graph.py respectively.

ASSIST

Last Update Date: March 23, 2022