MiTC

Introduction

MiLMo constructs a minority multilingual text classification dataset named MiTC which contains five languages, including Mongolian, Tibetan, Uyghur, Kazakh and Korean.

We also use MiLMo for the downstream experiment of text classification on MiTC.

Hugging Face

https://huggingface.co/datasets/CMLI-NLP/MiTC

Citation

Plain Text:
J. Deng, H. Shi, X. Yu, W. Bao, Y. Sun and X. Zhao, "MiLMo:Minority Multilingual Pre-Trained Language Model," 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Honolulu, Oahu, HI, USA, 2023, pp. 329-334, doi: 10.1109/SMC53992.2023.10393961.

BibTeX:

@INPROCEEDINGS{10393961,
  author={Deng, Junjie and Shi, Hanru and Yu, Xinhe and Bao, Wugedele and Sun, Yuan and Zhao, Xiaobing},
  booktitle={2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC)}, 
  title={MiLMo:Minority Multilingual Pre-Trained Language Model}, 
  year={2023},
  volume={},
  number={},
  pages={329-334},
  keywords={Soft sensors;Text categorization;Social sciences;Government;Data acquisition;Morphology;Data models;Multilingual;Pre-trained language model;Datasets;Word2vec},
  doi={10.1109/SMC53992.2023.10393961}}

Disclaimer

This dataset/model is for academic research purposes only. Prohibited for any commercial or unethical purposes.