Self-Supervised Graph Learning for Long-Tailed Cognitive Diagnosis

Authors: Shanshan Wang, Zhen Zeng, Xun Yang, Xingyi Zhang

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on real-world datasets show the effectiveness of our approach, especially on the students with much sparser interaction records.
Researcher Affiliation Academia 1Anhui University, He Fei, China 2Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, He Fei, China 3University of Science and Technology of China, He Fei, China
Pseudocode No The paper describes the methodology in text and uses figures (e.g., Figure 2) to illustrate the model, but it does not include a dedicated pseudocode block or algorithm listing.
Open Source Code Yes Our code is available at https://github.com/zeng-zhen/SCD.
Open Datasets Yes We conduct experiments on two real-world datasets: junyi 1 and ASSIST 2. 1https://pslcdatashop.web.cmu.edu/Dataset Info?dataset Id= 1198 2https://sites.google.com/site/assistmentsdata/home/20092010-assistment-data/skill-builder-data-20092010
Dataset Splits Yes To explore the effect of different sparse data on the experimental results, we divided the data set into different proportions. train:test 5:5 6:4 7:3 8:2 methods acc rmse acc50 rmse50 acc rmse acc50 rmse50 acc rmse acc50 rmse50 acc rmse acc50 rmse50 (Table 2 header)
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments.
Software Dependencies No The paper states "We implement our SCD with Py Torch." but does not specify a version number for PyTorch or any other software dependencies.
Experiment Setup Yes For each model we set the batch size to 256. As for graph-based models, i.e. RCD and SCD, we set the layers of the graph network to 2.