Official Code for paper "GMOCAT: A Graph-Enhanced Multi-Objective Method for Computerized Adaptive Testing" (KDD 2023)
Python, PyTorch, dgl
We have provided the preprocessed dataset assist2009
so that we can directly run experiments on IRT
with assist2009
.
To run GMOCAT, please run train_gcat.sh
.
- put raw data in
raw_data/
. - run
preprocessing.py
andconstruct_graphs.py
. - run
pretrain.sh
. - run the selection algorithms
train_gcat.sh
.
CAT baselines can be found in https://github.com/bigdata-ustc/CAT, https://github.com/bigdata-ustc/NCAT and https://github.com/arghosh/BOBCAT.