The source code for "A Simple Graph Contrastive Learning Framework for Short Text Classification"
Here we provide two datasets, snippets and mr, for reproducibility.
You can run the following the command.
python train.py --dataset 'snippets'
You can change the --dataset to mr to train another dataset.
We also provide the original datasets used in our work, along with the preprocess.py script to convert the raw data into the required target format.
Our related work can be found here: GIFT and MI-DELIGHT.
If you find our work can help your research, please cite our work!
@inproceedings{liu2022few,
title={Improved Graph Contrastive Learning for Short Text Classification},
author={Liu, Yonghao and Huang, Lan and Giunchiglia, Fausto and Feng, Xiaoyue and Guan, Renchu},
booktitle={Proceedings of the Thirty-Eighth Conference on Association for the Advancement of Artificial Intelligence (AAAI)},
year={2024}
}
@inproceedings{liu2025boosting,
title={Boosting Short Text Classification with Multi-Source Information Exploration and Dual-Level Contrastive Learning},
author={Liu, Yonghao and Li, Mengyu and Pang, Wei and Giunchiglia, Fausto and Huang, Lan and Feng, Xiaoyue and Guan, Renchu},
booktitle={Proceedings of the Thirty-Nineth Conference on Association for the Advancement of Artificial Intelligence (AAAI)},
year={2025}
}
@inproceedings{liu2025simple,
title={A Simple Graph Contrastive Learning Framework for Short Text Classification},
author={Liu, Yonghao and Giunchiglia, Fausto and Huang, Lan and Li, Ximing and Feng, Xiaoyue and Guan, Renchu},
booktitle={Proceedings of the Thirty-Nineth Conference on Association for the Advancement of Artificial Intelligence (AAAI)},
year={2025}
}
If you have any question, feel free to contact via email.