Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
text
Languages:
Indonesian
Size:
100K - 1M
License:
File size: 904 Bytes
5d59135 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
# "Clean" mC4
[[Original Dataset](https://github.com/allenai/allennlp/discussions/5265)]]
Indonesian mC4 with filtering from spam, NSFW, and Gambling. Big thanks to AllenAI for releasing Multilingual C4 🙏.
## Filtering Steps
1. Remove text from CCAligned domain (Why? CCAligned sometimes contains machine translated website.)
2. Find porn-y and gambling words, if it exceeds certain threshold it wont be included.
3. Train fasttext clasifier from (NSFW & Gambling) and (Indonesian CC-NEWS) text and classify text with it.
## Citations
```bibtex
@article{Raffel2020ExploringTL,
title={Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer},
author={Colin Raffel and Noam M. Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and W. Li and Peter J. Liu},
journal={ArXiv},
year={2020},
volume={abs/1910.10683}
}
```
|