Thank you very much for this work! It is highly interesting, and kudos for the effort made to open the code and datasets.
I am working a bit with the Moji Dataset, I would like to use the pre-encoded tweets you provide. I managed to run the pipeline from demog-text-removal, so I now have the raw content.
The issue is that we don’t have access to the split you used when preparing the data in download_data.sh. Could you provide the code you used? Or better, a second folder with the raw text aligned with the DeepMoji embeddings?
Dear Shauli,
Thank you very much for this work! It is highly interesting, and kudos for the effort made to open the code and datasets.
I am working a bit with the Moji Dataset, I would like to use the pre-encoded tweets you provide. I managed to run the pipeline from demog-text-removal, so I now have the raw content.
The issue is that we don’t have access to the split you used when preparing the data in download_data.sh. Could you provide the code you used? Or better, a second folder with the raw text aligned with the DeepMoji embeddings?
Best, Antoine